# AI-Powered Rails Upgrades with Ernesto Tagwerker: NextRails and the Future of Framework Modernization | The Ruby AI Podcast | Def Method

> Ernesto Tagwerker on AI-powered Rails upgrades, dual booting legacy apps, encoding expertise into AI skills, and how FastRuby is beating project estimates with one human and one agent.

[← All Episodes](https://www.defmethod.com/podcast)

Episode 22
•

July 07, 2026

# AI-Powered Rails Upgrades with Ernesto Tagwerker: NextRails and the Future of Framework Modernization

with Ernesto Tagwerker

![AI-Powered Rails Upgrades with Ernesto Tagwerker: NextRails and the Future of Framework Modernization](https://www.defmethod.com/assets/podcast/the_ruby_ai_podcast_episode_22-7d83bd604da0880584e885a649929c40c4d86b1eb57b6389c046b598f75677af.jpg)

## Listen to this episode

[Listen on Web](https://www.therubyaipodcast.com/2388930/episodes/19457740-ai-powered-rails-upgrades-with-ernesto-tagwerker-nextrails-and-the-future-of-framework-modernization)
[Spotify](https://open.spotify.com/episode/0cFoDw90LlKth3BRmKdZxB?si=add869ce194b496d)
[Apple Podcasts](https://podcasts.apple.com/us/podcast/ai-powered-rails-upgrades-with-ernesto-tagwerker/id1792967365?i=1000775803121)
[YouTube](https://youtu.be/ST-pf572bbY?si=srQ1dM8fn-ZpvhaP)

## About Ernesto Tagwerker

Ernesto Tagwerker is the founder of OmbuLabs and FastRuby, a team that has completed more than 60,000 development hours across eight years of Rails upgrades. He is the creator of NextRails and a longtime contributor to the Rails upgrade ecosystem.

## About This Episode

Rarely does a story about an AI tool actively resisting instructions end up being the most compelling argument for using that tool. When Claude initially pushed back against adding conditionals for dual booting large legacy Rails applications, the team at FastRuby had to essentially teach it their own hard-won expertise rather than letting it default to general Stack Overflow consensus.

Ernesto Tagwerker from OmbuLabs and FastRuby joins Valentino and Joe to unpack what that process actually looks like in practice. The conversation covers dual booting — running test suites against multiple Rails versions simultaneously — encoding human experience into AI-usable skills, and the shift toward outcome-based value rather than hourly billing.

After more than 60,000 development hours across eight years of Rails upgrades, FastRuby has started consistently beating their project estimates with one human and one AI agent matching two humans' output.

What does it actually mean to keep humans in the loop when AI handles ninety percent of implementation work? Ernesto argues that Ruby's readability makes it especially valuable precisely when humans need to oversee and debug AI-generated code.

## Full Transcript

Valentino Stoll (00:01.006)
Hey everybody, welcome to another episode of the Ruby AI Podcast. I am one of your hosts today, Valentino Stoll, and I'm joined by Joe.

Joe (00:08.399)
Hi, I'm I'm Joe. I'm the other guy. I'm just stepping over Valentino's lines now. nice to meet you all. It is a heat wave here in New York, and I think it is really just here to train us New Yorkers that are going to RubyConf to deal with the heat 'cause it's gonna be like a hundred and twenty five degrees in Vegas in a couple of weeks. I don't even know if I'm gonna leave the hotel.

Valentino Stoll (00:29.902)
I think they just hose you down with like slushies, right? Like

Joe (00:31.473)
Yeah. Yeah, that would be nice. You know, we were gonna roll out our our our new swag, our new hoodies. The the death method hoodies are are legendary. and there I was kind of of two minds about it. It seems really strange to to give people a hoodie in a hundred and twenty degree heat, but then again, if you never leave the the hotel, it's probably gonna be cold. So anyway. But either way, we didn't get off in time, so sorry. We're gonna have to get to you some other time.

Valentino Stoll (00:42.242)
Ha ha.

Valentino Stoll (00:54.616)
True. Yeah, you know they have they have those new like cut off like sun hoodies. I don't know if you've seen those. Yeah. I see golfers wear those a lot, you know.

Joe (01:02.664)
yeah, that's true. That's true. you know what? That's a good idea. Yeah, yeah, yeah. I'm gonna think about that. We can get a couple of new options. anyway, we've got somebody here with us. Why don't we introduce to everybody Ernesto, tag worker. Ern Ernesto, say hello.

Ernesto Tagwerker (01:20.371)
Yeah, hey everybody. So great to be here. I've been a listener and happy to join Joe and V for the first episode.

Joe (01:28.706)
All right. That's right. So we have had representative Ivan Boo Labs on before, and we went really, really deep into the weeds into AI, which was cool because especially at the time, you know,

a lot of people and I guess even still today, a lot of people say, Hey yeah, we do AI, which basically means we we use an API to to get Claude or ChatGPT to do things for us, which is not really the same thing as being like you know an ML engineer or you know or you know working with AI at a deeper level. but Ombu Labs definitely does. You want to tell us a little bit about that?

Valentino Stoll (02:00.917)
Mm-hmm.

Ernesto Tagwerker (02:10.703)
yeah, sure. so yeah, Amanda is our head of AI. She was in a previous episode and I'm sure you guys talked about machine learning, gen AI, LLMs, RAG, all that stuff. and at Umbu Labs what we do is basically custom AI solutions like yeah, predictive machine learning model, gen AI solutions, and the other side of our business is also like AI strategy for small businesses. ton of businesses here in Philadelphia are reaching out.

Joe (02:21.529)
Mm-hmm.

Ernesto Tagwerker (02:40.458)
Because you know, FOMO. Basically they know they need to use AI to make their processes more efficient, but they don't want to go about it, you know, just winging it. So usually like safety and compliance are top of mind for some of our clients. So they engage with us so we can guide them into what tools and what set of tools are best for them.

Joe (02:53.068)
Yeah.

Ernesto Tagwerker (03:02.158)
And on the Ruby side, we also run a product I service called fastruby.io. And on that one, we also are using AI and open sourcing some AI tooling to help companies make upgrades faster and get from all the way from Rails two three to eight.

Joe (03:22.324)
Yeah. So the open source contributions are awesome and and I've heard you speak in the past about Next Rails, which is which is a really cool dual boot gem so that you can sort of understand you know what's gonna break when you upgrade. so why don't you tell us a little bit a little bit about that and then you know tell us a little bit about some of the sort of next generation AI tools that you've been building.

Ernesto Tagwerker (03:49.061)
Yeah, sure. So NextRails is yeah, like you said, a really cool tool that you can use to dual boot any bundler based Ruby application. So it works for Rails, but it can also work for Sinatra or Sidekick. and we like to use it to basically test, you know, one version and another version very quickly by just flipping on and off like an environment variable. And it works very well from like development to test to production, even if you want to do like gradual deposit.

deployments and you don't want to do like one big bang deployment for Rails five two or something like that. then on top of that we started using Claude a lot recently. we actually kinda went in on Claude about three months ago and we said, okay, we need to build the tooling on top of Claude to use Next Rails. and it all came out as a basically as a rant that I had where if you ask Claude to upgrade

Rails from X to Y, it will make a case against dual booting. And I get it. You know, if you have like a monolith that's been around for more than 15 years, you have a ton of lines of code. Yes, there are gonna be like a many if-else statements that say if current Rails do this, if next Rails do this other thing. so Claude was like, no, I don't want to add this thing because it'll be like too many conditionals. And I was like,

from experience I know that I want the conditionals because it will make it easier when things get tough. so that's how three different sets of skills came to be is basically get clot to dual boot, get clot to

use the test suite as much as possible and then at the end of the upgrade flip the Rails defaults to use the last one, the the the most current ones. So that's a little bit about like the AI tooling we've been using. And then another part is like we built an automated AI assisted roadmap for anybody who wants to generate a roadmap from Rails two three two eight one. But that one hasn't gotten a ton of usage. So we're thinking about

Ernesto Tagwerker (06:02.88)
What to do with that next. For now, I think our focus is on better tooling for Claude or Copilot. so yeah, that's a l a little bit about what we've been doing lately.

Joe (06:13.936)
So it's interesting, and we were talking about this just before we started the show. first this this resistance, you know, comes from somewhere. you know, presumably a bunch of you know, stack overflow discussions about adding conditionals, right? Because, you know, L you know, the L L didn't come up with its own ideas about adding conditionals. So but it sounds like there's some kinda some element of you know, you have to lower the resistance.

And you know, and make sure it says, Hey, actually this is a good idea. Right. And so is that is that just added context, you know, or is it is it done through repetition? You know, how does that work?

Ernesto Tagwerker (06:53.642)
Yeah, I think we've only seen seen it work without conditionals with really small Rails applications. So our own experience is based on, you know, really big engineering organizations that need to upgrade Rails, like

Joe (07:00.102)
Mm-hmm.

Ernesto Tagwerker (07:08.918)
Power home remodeling was one of our clients in the past and they have a huge ERP that has, you know, a big team behind it. So upgrading their Rails application wasn't a trivial thing. And most of our clients have those sort of applications that are really big. so I find like when when you use AI, like the best thing is to combine it with some some human experience, to be honest. So you don't want AI to make those decisions for you. Like you said, it's basically like take

Joe (07:14.395)
Mm-hmm.

Ernesto Tagwerker (07:38.814)
Taking the averages from Stack Overflow, Reddit, you know, open source projects. So I see it as a way to encode our experience and our process into a skill that's like readable and easy for Claude or other Gen AI tools to to enter interpret and there's like no no gray area. It's like you dual boot and if you don't want to dual boot that that's cool. Just don't use this skill, you know.

Joe (07:41.937)
Yeah.

Joe (08:02.428)
Mm-hmm. Yeah.

Valentino Stoll (08:06.219)
Yeah, you know, I remember first seeing dual booting one of the Rails comps. from I I think maybe it was Shopify Folk. I d I don't really know. what was it, the bootloader gem or something like that? and it made so much sense, right? Yeah, boot boot. and it like you know, it's the best way to just s okay, like as you're going and, you know, making updates to your app.

Ernesto Tagwerker (08:21.239)
boot boot.

Joe (08:23.791)
Yeah.

Valentino Stoll (08:31.691)
you just like in the background run your whole test suite against the dual booted Rails app and then you can like catch bugs like that may be in the latest versions. Like it made a lot of sense. and so I I get why you're like focusing on that. to upgrade Rails 'cause it's like the easiest way to just see am I prepared to upgrade? Right? Which I think is like probably one of the biggest things when you're upgrading any app is like, you know, is this gonna cause any problems when I try and upgrade it or

Am I just gonna have this like massive PR at the end that's gonna like then be broken? Right. so I I'm curious, like did you take that approach like to like do incremental upgrades with this Next Rails? how how does that aspect of things like work within Clone?

Joe (09:03.417)
Mm-hmm.

Joe (09:11.245)
Ernesto Tagwerker (09:19.125)
yeah, yeah, we we see it also as a way to detect like backwards compatible changes as well. Like you need a conditional if it's not backwards compatible. But another thing that's cool about Claude is that you can write

code that's basically like monkey patches or little Rails shims that are like, okay, this is not gonna pollute the entire code base in you know the whole thing like if this every model that no longer extends like active record base. Now it ext extends like application record or whatever. Or like the Rails migrations now take brackets or something like that. now for s those things instead of like doing like if else is in like s a hundred and seventy files you can just basically write

A little rail shim that kind of plugs into the live the the libraries directory. And back in the day without Claude, I think that was like a little bit too complicated to do. So maybe we would we would default to like if-else statements. But now with Claude, it is quite good at generating those shims. and I know it's like not great, and I would definitely not advise people to start monkey patching everything. I'm just saying like for this particular case of like upgrading Rails, it's a really good fit.

Joe (10:36.525)
Yeah, I remember, you know, replacing active record base in, you know, many, many times by hand, right? And so that's that's not fun. I'm curious w either with the shims or with conditional logic, are you are you tearing those down at the end of the upgrade, or are you keeping some of them in? Or is it is it not all or nothing?

Ernesto Tagwerker (10:56.995)
No, no. No, at the end of it you you get rid of them. and usually like the best practice for a shim like that is to actually use like the if

If Rails version is current, do this, otherwise do this other thing. So even if you flip the switch, you know, set the environment variable in production to use like the latest Rails version, that shim should have logic in it to to not run in the wrong Rails version. so that's usually how we do it, you know, very defensively.

Joe (11:28.831)
Yeah. Right, but you're so how do you know 'cause you're putting these things in, are you putting in like how do you know what to clean up when it's over, right? 'Cause you presumably you're not gonna do it immediately. You wanna make sure that everything's running and, you know, have it in production maybe maybe I don't know, a week, you know, a month, and then you're like, Okay, good, we're we've made it. so how do you know what you're gonna rip out?

Ernesto Tagwerker (11:52.974)
Yeah, and that's actually one of the learnings we had as as soon as we started using the clot skill in like client applications. Juan, one of my teammates, suggested a cleanup job. So now I don't know if it's like a specific skill we have, but it might be like one workflow we have within the skills to clean up all these Rails upgrade related code out of the code base.

So yeah, we we are using Claude for for the cleanup as well. and I don't know about you, Joe, but we are seeing like really good speed ups on upgrades, like up to the point where you know, now it it is a reasonable

a reasonable case to say, we could have one human and one agent or one Claude basically working on an upgrade and get comparable results to maybe two humans working on an upgrade. I'd love to know what your experience has been using AI for for upgrades if you you know, 'cause you also run a a consultancy.

Joe (12:53.895)
Yeah. Yeah, we do. so we don't offer this as a service, but we obviously we upgrade Rails for for clients we're working on and yeah, the results I mean it's it's both it's both faster and it feels more sturdy. right, so our engineers are moving with more confidence. and pr presumably that's because or not presumably, I mean I've talked to them. It's because, you know, they have they do have Claude or they have

Ernesto Tagwerker (13:14.679)
Mm-hmm.

Ernesto Tagwerker (13:19.971)
Mm-hmm.

Joe (13:23.964)
you know, any of the other AI tools doing the majority of the work. So they spend much more of their brain power, you know, ensuring that what is happening is logical rather than, you know, typing everything out and e you know, spending time doing that. And so even if they're spending more wall clock time analyzing the results, what comes out is yeah, it's it's happening faster and it's happening with more confidence. The biggest thing I've seen Ernesto, and I maybe you see this too,

This year it's been pronounced. Last year it was not like this as much. all my engineers are beating their estimates every time. Which is a thing, you know, I've been r leading teams for like two decades and that's not a thing that has ever happened. We're just like we just famously underestimate all of our, you know, all the you know, the work that it takes, and now we're actually beating our estimates on a consistent basis, which has been really an interesting shift. I don't think that'll last, because pretty soon we're gonna get overconfident.

Ernesto Tagwerker (14:00.226)
Mm.

Valentino Stoll (14:04.118)
Mm-hmm.

Ernesto Tagwerker (14:05.835)
Mm-hmm.

Joe (14:21.359)
and then we'll start underestimating again. But but for this year at least, it's been this interesting thing where it's like, well, I think, you know, it's probably gonna take three months and it's just from lived experience. It's like, yeah, I know this thing'll go fast, but not that fast, and it's not gonna be that good, and we're gonna have to go back and clean this stuff up. And you know, that's true to an extent, but it's also true that we're you know, we're overshooting it a little bit.

Valentino Stoll (14:44.28)
That's great to hear.

Ernesto Tagwerker (14:44.33)
Yeah, yeah. And I think I I've seen that as well, like on the on the upgrade side. But

Joe (14:46.374)
Yeah. You have? Yeah.

Ernesto Tagwerker (14:50.902)
Then on the custom AI solution side, I haven't seen that so much. And I think like it's just us being like overly optimistic that we can deliver like a non-deterministic solution to this client in you know, this amount of time, and then there's always like things that come up. and part of it is like managing expectations, right? Like people who are getting custom AI solutions expect a deterministic solution. And sometimes there is like, you know, the part of non-determinism that is like, hey

Joe (15:03.26)
Mm.

Joe (15:11.525)
Yeah.

Ernesto Tagwerker (15:20.855)
you know, for this particular use case we wanted to use an LLM and it's gonna generate content and sometimes, you know, there's gonna be something that's off and it can't be a hundred percent, but it can be close to that, right?

Valentino Stoll (15:30.958)
That makes me think.

Joe (15:33.38)
It could be close, yeah. I have this I have this experience with my assistant who's using Claude Cowork every day. And we have this process internally of this this sales engine that that we built. and so she has a part in it which is like she's using Claude Cowork to basically, you know, search through LinkedIn and look for these people and add them to this particular list. And and it keeps getting it like ninety percent right and it's basically it's like there's other parts where it just

Valentino Stoll (15:33.954)
Yeah.

Joe (16:03.385)
And it just kinda it doesn't fall over. It just puts you know, puts like LinkedIn URLs that are non existent or you know and she keeps say like I keep coming back to her and being like, Hey, it's better 'cause I'm like, hey, this is really good. Like ninety percent is is correct, ten percent isn't. And she keeps being like, Well, why is it doing this? Why? And I'm like, Yeah, this is yeah, this is kind of a learning experience. You're it's not anything you're doing. This is just non deterministic software. It's gonna do it right a lot and it's gonna mess up sometimes and that's where we come in.

Valentino Stoll (16:18.646)
Yeah.

Valentino Stoll (16:26.762)
Yeah.

Joe (16:29.73)
Sorry V, go ahead. I know. I know. Yeah, no it is. I look at it and I'm like this is just like LinkedIn slash in slash John. Like that's not real. That's not that's not a real thing.

Valentino Stoll (16:29.858)
The edge cases are really funny. it doesn't matter the task, like it's al it always entertains me. Yeah.

Valentino Stoll (16:42.446)
Yeah, I was gonna say, you know, I'm curious, you know, I haven't done a I I automate all my Rails upgrades at this point because they're all just personal projects, right? So like in the background, I'll just be like, hey, go upgrade Rails, right? and then I realize, things are broken and I fix them. 'cause they're just for myself. but like I wonder like, you know, if you had to draw a boundary, like what should the the AI upgrading agent like

be allowed to automatically do versus like where does the human take the wheel? Right? Like I'm curious like where you're seeing kind of like the things fall apart versus like what's for sure, yeah, I trust this thing is gonna just do this specific thing.

Ernesto Tagwerker (17:28.23)
Yeah, that's a good one. basically the deprecation warnings, I would say those can be like on autopilot. Like Rails, I mean, especially if you're using like s modern versions of Rails, like the deprecation warnings are way more complete and present and there are not that many things that are undocumented. So addressing a deprecation warning I would just say like yeah, go and do it and maybe submit like a pull request per deprecation warning if you wanna review it

And have a human in the loop there. for anything that's front end, I would

want a human to review it and to to QA it thoroughly before it ships to production. because I I don't know about you guys, but I have noticed that Claude and you know Copilot and all those things, they're great. But then at some point they're like, okay, this is done and then you look at it and it's like, this looks terrible on the screen. And you know, the the robot might be like, No, no, this looks fine. It's it's fine. So for anything that's like full client facing

Maybe like in the UI, I would want to have a human in there. but yeah, I I would make those those big distinctions on on that front.

Joe (18:48.171)
Yeah, I mean one thing that I f I find routinely that the AI thinks is fine. This is Ruby specific. I I mean it it might be like this in other languages, but a couple of things that it routinely thinks is fine that is not fine. really long classes in Ruby. It kinda doesn't seem to sense, no matter how many different ways I prompt it, that

You you don't want to add another method here. You know, like there's there's already enough methods. and and it does not seem to to really do much object orientation or object individuation beyond the the f what the framework itself gives you. Right. So unless I tell it, hey, we're going we're going VP or we're going with a separate service library or we're going with you know it doesn't it doesn't tend to do it. I don't know if that's matched your experience or or yours v.

Ernesto Tagwerker (19:43.182)
Yeah, I I don't know. I think one one area where I wish it added more value and I'm sure Joe has like a lot of insights on this, is the writing tests or increasing tests before an upgrade. I know, you know, you worked on Phoenix deaf method and I don't know, I just would love to see a tool or a set of like skills that basically

feeds off like data from production and it writes test suite that are based on like reality, which is production usage. and of course like for V that is easy for like a side project or a fun project because you can just give it access to production and it's fine. for client projects that are in like in regulated fields, which you know are many of our clients, we would have to have some sort of like layer to make sure like any sort of like PII or sensitive information never makes it to

the Gen I Gen AI engine. but I I can see like a future where it's like, okay, you have a robot that is slowly and gradually increasing code coverage for some of these actions in production that are getting called over and over again and have like very little code coverage. but I love to to live in that future where it's like, I have an agent that's submitting pull requests and the pull requests are not based on some hallucination or just increasing code coverage for this class. They're actually

Joe (20:58.866)
Mm-hmm.

Ernesto Tagwerker (21:12.83)
based on real usage, could be like APM A APM data too, to say like, these actions are called all the time and you don't have tests for it, so I'm gonna submit a pull request to add a test for that.

Joe (21:27.473)
Yeah. I mean one thing we learned in building Phoenix is that the the differences between application test suites are are vast. even in in seemingly very even even on code bases with seemingly very similar characteristics. I love the idea of of sort of a sanitizer layer that pulls production data, sanitizes it and then sends it to an L L to to

build tests. I don't think frankly we're at the place where that can be sassafied, right right, where it could be like, okay, this let's build a one size fits most solution here. but I do think it could be a great service that's offered by companies like yours, Ernesto.

So go ahead and do that for us. Okay.

Ernesto Tagwerker (22:23.511)
Okay, okay. I'll I'll think about it. I mean the problem with AI these days is that all of my side projects are doable. and of course it's like combined with my w wishful thinking

Valentino Stoll (22:25.262)
Ha ha ha.

Joe (22:34.064)
Yeah. Now now you're speaking Valentino's language. Yeah.

Valentino Stoll (22:34.166)
Yeah.

Valentino Stoll (22:40.238)
I know. It's it's you know, I I just wish I worked for one of these companies and had unlimited tokens and then I would just like, you know, do them all at once. But I ha I have budgeting concerns. but I'm curious like what what you're talking about, it sounds like and this Rails Next project too. it reminds me a lot of the the GitHub scientist gem. I'm not sure if you're familiar with that.

Joe (22:50.726)
Yeah.

Joe (22:54.863)
I know.

Valentino Stoll (23:07.736)
But it was like a way to test like critical code pass to like basically try to execute code in two different things or in two different blocks and then you know measure the outputs to make sure that they were exactly the same or the outcomes. it's not exactly like a hundred percent, right? Like 'cause how do you measure side effects or data like d database database transactions and things like that? but

It reminds me a lot of that. So I'm curious, like, have you thought about or what do you think about like that style of like reinforcement learning-ish processes, right? Where, you know, we have these two paths that we want like an app to take and we just want to make sure that they do the same thing eventually. Right? and it seems like that's kind of the path you're taking with NextRails. Like, is there a reasonable next step to that?

Ernesto Tagwerker (23:58.43)
Yeah.

Valentino Stoll (24:05.185)
where the agent can do a step function to automatically handle all of that, or it do you not see that future?

Ernesto Tagwerker (24:14.534)
Yeah, I don't I don't know. I I like the idea of keeping libraries small and like very focused on certain things and I like you said, there's like the Ruby Scientist gem. I haven't really used it, but yeah, it is really cool if you want to test like, maybe this particular query could be more performant if I structure this active record query like in a different way.

But for NextRails, I think that's the scope that it has is okay and and it solves the problem really well.

And I think where adds the most value is with gradual deployments where you can basically set it up to basically if you have like Kubernetes infrastructure, you can have like ten percent of the pods run on the next version of Rails and then the rest of the traffic use what what you've been using for so far. and that way you can keep track of like the error rates in your exception tracking service and you know, if there's like an elevated number of errors or new errors happening, it's like you know that ten percent of the traffic is like

causing those. But yeah, for for experimenting, I think probably like the Ruby Scientist Gym is still like the the best thing to do.

Valentino Stoll (25:33.773)
Yeah, you know, I s I was kind of like setting you up a little bit, because I my my real question is you know, how do you stop like getting too excited about the thing that you're making, right? Where all of a sudden you have like a million features that you could potentially do and realistically do if you threw enough like tokens at it, right? how do you like personally like pair that back and keep things well scoped?

Right, like you're mentioning, you know, you kept it very well defined so it did this very specific thing that is helping you. Like where do you see where are you value shaping the excitement and pairing it back to keep things focused, right, for for what you're doing?

Ernesto Tagwerker (26:13.914)
Yeah.

Yeah, that's that's a a good one 'cause again with AI you have no boundaries anymore. It's like and I'm I'm the owner of FastRuby and Umbu, so I don't have anybody saying like, no, you can only use these amount of tokens, so that's I should probably have that. maybe Amanda should do that for me. but yeah, I think the the problem that we have on our end is that we've been doing engineering as marketing for so long

that we have so many tools now. like we maintain Ruby Critic. we have this idea to basically have a back-end server to basically send your Ruby critic reports to. I haven't done it yet, but I might have like Claude do it for me and basically offer it for free for like open source projects.

We also have RailsBump.org that we inherited from Manuel, a software engineer from Germany that basically tests compatibility between like so many Rails versions and so many Ruby gems. It's basically kind of like RubyGems.org, but it adds compatibility with different versions of Rails and it has like different check techniques and

We have like the automated roadmap that generates an action plan. We have the cloud scale. And my problem right now is that they're not talking to each other. So one of our goals for this quarter is to basically get Cloud to talk to Rails Bump, get the automated roadmap to talk to rake Rails Bump. Next Rails itself has a way to calculate compatibility, but I don't trust it as much as RailsBump.org. So should we even have that in the next?

Ernesto Tagwerker (27:59.497)
Next Rel's gym, I don't know. Those are the questions I have right now. It's like we have so much code right now to figure out compatibility and the upgrade path and integration with AI tools that I'm having a really hard time like integrating it all to not do the same thing in three different places.

Joe (28:20.49)
Hey, do you get I've been reading about some people experiencing this. Are you getting a ton of new pull requests that are almost entirely AI generated?

Ernesto Tagwerker (28:36.022)
Yeah, yeah. I we are

Joe (28:38.065)
And is that is that helpful or is it annoying?

Ernesto Tagwerker (28:41.206)
I I have actually seen them and I have found them quite helpful most of the time. There's like one out of ten that's kinda like a hallucination and it's like please don't do that, go away or review your work. But no, one thing that we have been seeing on the consulting space and I don't know if you're seeing this, but it's basically this AI generated scams that reach out to build some sort of tool for you and then it's like an elaborate scheme to

Joe (28:45.893)
good.

Ernesto Tagwerker (29:11.081)
I don't know, get you to leak data to them. I don't follow them. They just we kinda have a we kinda have like an ear or an eye for that and it's like, okay, this this project description looks like way too good for to be like a real project description. So I'm like, it's probably a robot, some sort of scam. Do you do you get those two?

Joe (29:22.718)
Mm-hmm.

Joe (29:27.133)
Ha ha ha

Yeah. I have and I've read about them too and they're they're they're proliferating all over GitHub. there was some report on Hacker News, which Hacker News is where you go to find all the people complaining about the AI pull requests. So it's nice to hear you t say actually they're they're nice. They're usually helpful. but yeah, I read some report that some some I'm gonna forget the no the percentage, but some significant percentage of open source GitHub libraries right now are

scams that are trying to you know steal your data. So be careful out there folks. Yeah.

Valentino Stoll (30:03.362)
You know, they're they're getting pretty progressive. I saw I saw one lately in the Ruby AI Discord. Somebody posted a screenshot of them receiving money for this free deal, for having signed up for some crypto thing. And they're like, look, I just like got, you know, fifteen hundred dollars like transferred to my bank account. and you know so I f I was just having fun and I went through and like tried to sign up for the process and you get the final stage is like, okay, it's like some like weird casino in virtual casino in Carraso.

Joe (30:16.955)
Mm.

Valentino Stoll (30:33.536)
I don't know how you pronounce that. right. But like you know, in order to get your account like verified, you have to pay $200 so that you can, you know, verify your bank is like legit and you who you are. That's how they like framed it. And I'm just like, man, what a s what a great scam and elaborate, right? Like, is they have cause there was more than one person posting this. Like, yeah, me too. Like I got this thing, right? Like, and I was just like, Wow, like it's really getting elaborate. Like they had

Joe (30:34.865)
Yeah.

Joe (30:47.632)
Yeah.

Yeah.

Valentino Stoll (31:00.46)
Clearly like just an AI setup website, like that was had the sign up flow and it even had like an AI like customer a service yeah. Well they had like a a customer service agent set up, right? Whereas a chat bot. So you I had problems like setting up my account, like what is like, you know? And so it like s made it seem more and more legit and I'm just like wow, like somebody really spent a lot of time on this thing, like

Joe (31:05.904)
With AI setup Discord people, right? Yeah.

Joe (31:11.684)
Yeah.

Joe (31:16.186)
yeah, that's good.

Joe (31:23.428)
Now I think a really smart thing to do, not that I am purveyor of any of these scams, but you would actually want it to work for a few prominent people. Right? Like, you know, it actually no, I actually did get fifteen hundred dollars, right? That it worked like three people that you know, and then it just scams, you know, the other thousand people that try it.

Valentino Stoll (31:34.924)
Alright.

Valentino Stoll (31:39.95)
Well that you know, that's that whole you know, the Facebook thing where they're like, look, if you enter my you know, it has a special thing where you enter your password and it automatically replaces it with asterisks, right? And like a whole bunch of people comment, yeah, me too, look and it like just the asterisk

Joe (31:51.351)
yeah. Right, right.

Yeah.

Ernesto Tagwerker (32:00.073)
Yeah, you gotta be careful out there. So many scams and I don't know if you guys have seen the video of like the video, even like in video calls where the the person is like, Okay, now I need you to put like your your fingers in front of your face and it's like, like this and they put the fingers like you know, not on not in up there. yeah, you d you can't even trust video these days or video calls.

Valentino Stoll (32:00.257)
Anyway.

Valentino Stoll (32:11.886)
Mm-hmm.

Joe (32:14.828)
Mm.

Joe (32:27.158)
not. You shouldn't even trust this video call. Definitely shouldn't trust this podcast.

Valentino Stoll (32:28.974)
You know, I've I've also heard that of of interviews where companies hiring now get a lot of people interviewing virtually that are not people or are not who they say they are.

Ernesto Tagwerker (32:30.826)
Ha ha

Joe (32:43.308)
That's freaky. And that actually leads me to my next question, Ernesto, which is you had mentioned Amanda, who we've seen on the show, we think, and and I think in reality also. So I guess we're gonna we're gonna rule her out, she's a real person. You mentioned Juan. is that an actual human being?

Ernesto Tagwerker (33:00.591)
Yes, yes. We every picture we have on our website is a real human. I have actually met all humans in my team except one, Francois in South Africa, but I also believe he's a human and not a cat, yeah.

Joe (33:02.141)
Are you sure?

Joe (33:15.338)
Mm-hmm. Well, you believe it. So Francoise still maybe though. And and you've got at least according to you, some actual humans. Okay. Alright.

Ernesto Tagwerker (33:26.799)
Yeah. Yeah. Yeah. But no, it it is it is like getting harder and harder to filter and to do security checks on people that you hire remotely. So yeah, I I think that's that's one of the the the struggles we have.

But yeah, one of the things we we recently launched and this came out of like conversations I had with clients is we launched staff augments staff augmentation services. because my conversation with clients would be like, do you guys do like feature work? And I'm like, Yeah, we do more than Rails and Ruby upgrades. Like we would be happy to work on your product roadmap. And I had that conversation so many times that I was like, Okay, maybe it's not clear in our branding that Fast Ruby can also do product

product development. So for many years we focused on Ruby and Rails upgrades, but now that AI generated code is like more and more present, we also see like a ton of pull requests. So it does feel like the workload is basically shifting towards like code review and QA and rework. So this is like our idea to say okay you need a human in the loop to basically help your team either with code reviews or product development

development and you need them to be real humans that you can point to when they mess up. They're humans. They've been working with us for years. So now you can hire them to do product development for you or assist with code reviews or anything you won't need a human for.

Joe (35:04.842)
You know, it's good to know. of course you know being a consultancy owner myself, I mean I look at it and I'm immediately like, okay, yeah, obviously that is that's a that's a leader into the organization and then you you can do product development and things like that. But others may not see it that way, and so it's good to make it it's good to make it obvious and apparent because obviously the team that can, you know, provide a steady pair of hands and upgrade your your your Rails application is is likely gonna be the team that you want to steward

some of your product roadmap. I saw and it reminds me because I saw this post from you a few weeks ago that your ideal client has a table a stable test suite, knows where tech debt lives, wants slow and steady progress. Where do you find these companies? Because I've been running I've been running a consultancy for twelve years and I don't know if I've maybe I've stumbled into a couple, but that's not really the that's not doesn't tend to be the work that we get.

Ernesto Tagwerker (36:02.892)
Yeah, I think the test suite is the the main one, right? Like I think a lot of companies, especially startups or or companies that started as startups and they just w needed to move fast, they never stop to write the test suite or to keep it up to date or even to upgrade Rails and Ruby, so that's why they end up reaching out to us. the good thing is like when we work on our Ruby on Rails upgrade, we kinda get the a taste of what it is working with this client. And, you know, like everything, they're good clients, great

Great clients and bad clients. so it is a funny fact, like in the past year or so we've had way more repeat customers than new customers, which is a new thing for us. L

last year and maybe like two years ago it was more like, okay, new projects, new clients, new upgrade work. but now it's with AI shifting things and with people using AI for upgrades more, I feel like only the people that value like, you know, the human touch of working with us are coming back and saying, like, I actually had a good experience working with Fast Ruby, so I'm gonna reach out and work with them again.

But yeah, it it is hard. So that's why I think we invest so much in like technical debt and code quality tooling, like Ruby Critic, Skunk, Rails Stats, Next Rails, Rails Bump. All these tools are available usually for free or free as an open source. so anybody who has played with these tools and cares a little bit about code quality has probably heard or used some of our our open source lab libraries. that's

Usually like a good vector to find good clients. and sometimes it's it's hard to find good clients, but you do end up finding like good engineers within like a poor culture or poor engineering culture. so then you basically have to talk to them as a system to try to get like upper management and non-technical decision makers to say, okay, it does pay off to have like

Ernesto Tagwerker (38:10.432)
good code quality guidelines so that not just humans but AI are writing code that's gonna be like easier to maintain in the long term.

Joe (38:17.885)
Mm.

Yeah, I I can I I can understand your point. And I I also would say that we've had really good clients that don't have stable test suites, right? They're that's not really what we're after. They they maybe like you said, maybe you know, sometimes teams inherit, you know, code that's crazy. and sometimes t you know, teams move really fast in search of the

you know, what they're after. I had an engineer a long time, he was a great engineer, and you know, any time we would talk to people, 'cause a lot of times teams they they can get a little self conscious about

Were 'cause they kinda know where the bodies are buried and they're like, I wish it wasn't like this, right? And you know, he always would say, You know what, the the the greatest marker of your success is that you're here talking to us, right? Because, you know, you you did what you needed to do to build a successful product that's successful enough to pay the people that are working on it, you know, to pay other people, you know, drive a business, you know, to pay us to come in. and so we always want to be really respectful of that.

Anyway, but I'll get off my soapbox. That but that is you know, that's an interesting way of kind of framing it. When you're looking at these engagements, are you are you doing like a set like, okay, we're going from, you know, Rails, you know, five to Rails you know, eight, one, so y we'll just quote you a price, or are you kinda going time materials based?

Ernesto Tagwerker (39:49.648)
yeah, we usually quote

a price, like we like to do like ballpark estimates, like worst case, best case scenario between X and Y months to get you there. and of course, you know, the the more information we have, the better the estimates. So sometimes when we're having like an initial sales call, we will ask our customers or potential customers to send over stats. and stats are basically like, Yeah, what's your code coverage percentage as reported by SimpleCov? what are your rails stats?

As reported by our gem. Our gem kind of integrated with bundler stats, bundler stats, that basically gives you an idea of like the dependencies they have and how hard it'll it's gonna be to upgrade. and we have like historical database with more than 60,000 dev hours invested over the past eight years on upgrades, all the way from 2.3 to 8.1. So we kind of use comparable data to give our potential customers like ballpark estimates and then

Joe (40:33.379)
Mm.

Ernesto Tagwerker (40:50.832)
Then they make a decision there or sometimes they get a roadmap that's basically like a twelve thousand dollar audit where we invest about two weeks in their code base learning more about it and seeing kind of like where the bodies are buried and and stuff like that. And also like to complement the other

Joe (41:05.656)
Right.

Ernesto Tagwerker (41:11.43)
question, you know, or the the other answer I gave about like how to find like the best customers, like I'm still trying to figure it out. I just wanna make it clear. Like I don't have a solution for that. Like I post to LinkedIn to try to find them for sure. But

Joe (41:24.397)
Yeah.

Ernesto Tagwerker (41:26.2)
In the sales process, I can quickly tell when some clients are not gonna be great clients. And it's usually like the you know, the people that care way too much about the money and the budget and the hourly rate and they're just wanna get like three or four bids from deaf method, fast Ruby, test double, and they're like, Okay, cool, we have the numbers, okay, let's go with this one.

Joe (41:32.217)
Mm-hmm.

Joe (41:43.787)
Mm-hmm.

Joe (41:50.336)
Well, if they care about the money they're they're picking the wrong three. But I know I know exactly what you're talking about, yeah.

Ernesto Tagwerker (41:53.563)
Yeah. Yeah, yeah. But you know, they they don't really care about the craft and you can tell in the conversation, they just want the number. So I mean I still send the number, but I also like can write down in our CRM it's like, okay, this is like a ten percent likely to convert lead, you know.

Valentino Stoll (41:57.133)
Yeah.

Valentino Stoll (42:13.976)
Yeah, I'm curious if you like have found a like metrics template to translate to business owners, right? Okay, if you apply this, right, you'll you'll see this stability which has this kind of customer impact or outcome, right? like do you do you have anything that you've like been ruminating about, like where you can like sh like help them calculate numbers as like, okay, if you invest

This and this, right? Like you'll see this kind of outcome.

Ernesto Tagwerker (42:47.95)
I mean it sounds like a really cool really, really cool side project that V should work on. And I would be happy to give you tokens for you to just go wild with something like this.

Valentino Stoll (42:53.816)
Yeah.

Joe (42:54.775)
Yeah, that's nice, I know. Yeah. Yeah.

Valentino Stoll (42:56.994)
Yeah, okay. Alright. Hey, I'll take you up on that.

Ernesto Tagwerker (43:01.516)
I mean that sounds amazing and no, I don't have it and I should I should have. but yeah, the some clients are still, you know, trying to understand why a test suite or an automated test suite is a is a good idea.

Valentino Stoll (43:18.082)
Yeah, I mean I think I think about this too, even just for AI spend, right? Like, okay, if if you as a business owner, right, you spend, I don't know, fifty thousand dollars in a month, right, for AI tokens. Like, what is that outcome? Like I'm sure like you'd have a bunch of artifacts, right? But like it's gotta be more than the artifacts, right? Like what what is it generating for you as a business, right? Like, I don't know that there is that, right? Like which

Joe (43:43.966)
Think the I think the answer for every business owner is not enough. That's mostly what I hear from my clients.

Valentino Stoll (43:47.021)
Right, not enough. I mean, maybe if they get rid of an employee, like they they save some money on that. But like I feel like I don't know how much.

Joe (43:55.014)
Yeah, well that's that's a good that's actually I think you know, like it or not, I'm having conversations with with clients and I I spoke with somebody just last week about this where, you know, he's high up, he's at executive level and you know, the company got a huge investment from a PE firm and the you know, the question very quickly became, Okay, how many roles are we eliminating with the all the AI that we're putting in here? And I

I do think that that's the wrong question to be asking, but you know, it's not up to me. Right. Like I think I think the question to be asking is what what more can we do? And I think that's what you're getting at too, Valentino. Like, hey, what what's the what's that return on investment so that we can do more rather than, hey, where where can we cut corners and save?

Valentino Stoll (44:27.116)
Right.

Valentino Stoll (44:40.482)
Yeah, 'cause I I see this all the time. I was in consultancy for a long time. And like it's always like, okay, it's very layered, right? Like, yeah, you can go and you can like upgrade a Rails app, you know, but then like that's not really necessarily addressing any of the tech debt that's there, right? It's just like one aspect of the tech debt, right? And so like tech debt is like a very layered problem and like ha you have to balance that as a as a consultant for your clients, like, okay, you know

Joe (44:56.444)
Mm-hmm. yeah.

Valentino Stoll (45:09.378)
How much tech debt is worth addressing, upgrades included, versus like, okay, we have new features and we have bugs and things like this that you're helping them like realize and balance, right? because I imagine you want more more of that stipend, right, to help see them see more value out of you as a consultancy, right? and so I wonder, like Ernesto and Joe, right, like as these kind of consultancies, like where do you see that layering fit best, right? Like is

Joe (45:13.359)
Mm-hmm.

Joe (45:28.251)
Yeah.

Valentino Stoll (45:38.976)
Is an upgrade an easy entry point to just like helping address like the management of those concerns? Like where is AI like providing the most leverage outside of just like automating some of these low hanging fruit, right?

Ernesto Tagwerker (45:52.469)
Yeah, and what I like about the Rails or Ruby upgrade work is that the finish line is very clear. It's like it's in production and it's running the target Ruby and Rails versions we talked about and we delivered it on time and on budget. I love that. The problem I see these days is that people, especially like non technical folks, are looking at the hourly rate way too much.

And it's in this day and age of AI, the hourly rate doesn't really mean much. It's like the same of using lines of code to measure developer productivity, you know? It's like it's a waste of time to say like, line we generated all these lines of code with all these tokens and all these people.

doesn't really say anything about value. So like time and material is you know, soon a lot of these folks are gonna have to learn that the hourly rate is like a terrible way to make a decision. Cause like you could get like twenty hours of death method and their hourly rate, or you could get forty hours at from some other shop that's like way less expensive like per hour. But at the end of the day you care about value and you care about

you know, keeping the engineers and making sure that they're delivering value. So it is very hard these days to say, yeah, it is this much per hour and you get X amount of value. So I wish there was actually like an industry standard for value per hour in the consultancy space. But right now, yeah, we don't have it.

Valentino Stoll (47:27.007)
Yeah.

Valentino Stoll (47:32.164)
I've I've got a great I great idea, great great solution to this problem. Ready? We're gonna solve it all right now. just everything is tokens. Right? So like the you know Ombolab's tokens, right, which represent the hours that are put in, right, are more expensive than open AIs because you can do things that it can't, right? If everything's a token, the people stop looking at hours, right? And it's just like okay, you know, as as you're operating LLMs, like

Joe (47:32.663)
I think we've Yeah. Yeah.

How good?

Ernesto Tagwerker (47:37.599)
Go for it.

Joe (47:39.919)
Everything is tokens. Yeah.

Joe (47:53.135)
Mm-hmm.

Joe (47:57.155)
yeah.

Valentino Stoll (48:01.827)
You're paying for intelligence levels, right? Some intelligence is more than others, right? Your you know, your consultancy's employees are specific tokens of intelligence, right? That are gonna outperform any AI agent that they can get, any LLM, right? and so like I wonder if that's maybe the answer is just like, okay, let's bucket humans in the same AI worker category and just say, All right, you're gonna do get the omni omni labs you know, model.

Joe (48:05.551)
Mm-hmm.

Joe (48:23.735)
Yeah.

Valentino Stoll (48:30.743)
Right. And like th these are the our tokens token rates, you know.

Joe (48:34.015)
Right. It could be I d depending on who it is, I mean, I'm just thinking about getting that argument passed to CFO. But maybe just thinking about people in, you know, in units, that might be the way to go.

Ernesto Tagwerker (48:35.269)
Yeah.

Ernesto Tagwerker (48:48.016)
Yeah, well I think there is like a a point for like a new metric that's tokens is a part of it. For but for me it's like experience times tokens or exp experience combines with tokens. Let's say X tokens is like the value you get from these engineers. And then I don't know, may maybe that is the metric that we need to get to.

Valentino Stoll (49:11.299)
Yeah, 'cause you think if you're hiring a law firm, right, like the senior executive lawyers, right, are gonna have like some astronomical hourly rate compared to like some junior and like, you know, if you're hiring them and have on like some kind of retainer, you're gonna say, Hey, like, give them more hours, right? Like, sure it might take longer and they might do worse worse work, but like I just don't wanna spend that much, right? Like And so, I don't know if you can apply the same thing and still uphold like the quality of work that outputs, right?

But I wonder if there's s something there, right, where you could just say, Okay, yeah, this this level of person on our team is gonna be like an extra bump, right, in in the allotment of units.

Joe (49:51.221)
Right, in tokens.

Ernesto Tagwerker (49:55.834)
But yeah, I I also agree with Joe that getting the CFO to buy into this whole idea is gonna be the hardest part.

Joe (50:03.659)
Well, so Valentino brought this up before the show, and I we've got a few minutes, and I'd I would love to just get your take on it, Ernesto, which is why now that we have AI and it can do any it can do any language, it can work on any framework, why stick with Ruby? Right? Why why recommend it? Why even upgrade it? You know, why not just say, well.

You know, you're all the way behind at version five. Just blow it up, just start over with you know, Python application or something like that. You know, all the AIs love Python, so why not just use that?

Ernesto Tagwerker (50:44.152)
Yeah. no that's that's a good question. And we still choose Ruby for a lot of our custom AI projects. We might not use a hundred percent Ruby for that particular solution. It might be like a combination of Ruby and Python. but at the end of the day, I think you know, Ruby's still more readable than Python and is still, you know, focused on you know, developer happiness. And

We're not quite there where it's like we're just gonna let the agents run everything in production and fix books and yada yada.

I know V might be close to that, but you know, there are a lot of companies out there that still need a human in the loop that's basically approving the changes. So if that human has experience with Ruby and can go in and like figure out like what the AI is trying to do, and many times the AI AI is trying to do something that's like way too complicated for something that doesn't need to be that way.

I still want humans that know Ruby and can go in and like debug and like sort out what what's what's the problem. and basically th driving Gen AI tools to to write better Ruby and more sustainable Ruby. but I do I I I know you guys were talking about like Ruby Central in a past episode and

I like the way it's going now with Ruby Central and I know like they're they have an idea to be more present in the AI world and as a company we want to somehow support that. I don't know how, but at some point we want to c collaborate and like contribute to the community so that we can make Ruby more AI friendly. and if anybody has ideas on how we can do that, we're always happy to to entertain them. They can reach us like on Blue Sky

Joe (52:24.44)
Mm.

Ernesto Tagwerker (52:41.743)
LinkedIn and but for now I I like the initiatives that are happening at Ruby Central and I am hoping they can ba make Ruby like a better citizen in this AI world.

Joe (52:53.703)
I'm not sure I agree with you on on the Ruby Central side, but you know, we're you know, we're a supporter and and I definitely want to see Ruby Central succeed. to me it seems like they're they're trying to spread out in too many directions and what they really are is a very small group of you know unpaid volunteers. And I think there's there's an engineer's inclination, like I was talking about before, to overestimate what

what any group of us can do. And I unfortunately I think that's what's happening here. I also think that the the AI thing, I mean, I'll I'm gonna reserve I'm I'm actually gonna reserve judgment on that because I do think it's interesting and I'll be there at RubyConf to talk about any of these initiatives. I think that's what they're using, right? They're using the conference to to get to gather some like minded folks and talk about it. So I'll be interested to hear you know what they have to say on it. but I mean to me

And I I I want them to keep putting on a good conference. then to me that's the most important thing. and then, you know, I think the second most important thing is is maybe supporting conferences that are, you know, regional, which they've done for a very long time. and to me, you know, w we build software that is great at one thing. And if they were great at that one thing, and they have been, right? If they continue to be great at that one thing, I'm happy. All the other stuff I think is gravy.

Valentino Stoll (54:16.185)
Yeah, you know, I I I see it as a marketing problem, right, with Ruby and AI, right? Is like ha you know, it's the Coca-Cola method, right? Like you see everybody drinking a Coke and you want one, right? Like and I feel like that's kind of what we're seeing with Python and TypeScript is it's just like, okay, like that's what's used, everybody they see other people using it, like

you're starting up a project and like the AIs are using that, right? and I would love to see, you know, some of these organizations like Ruby Central like focus on that, start trying to solve that particular problem, right? Like Joe mentions like conferences, like that's one aspect, right, of many. And I think like the whole Rails world thing that like Amanda has been focusing on, right? Like fantastic, right? Like

getting Rails out there and advertise worldwide, right? Like is like an aspect that has like I think seen positive results from people, right? And I think the more that we could get that, like the easier it's gonna be to solve like the tooling layer, which I think is more like, you know, Ruby core team, right? Like dump money into that, right? 'Cause like they're really solving those problems. I would I would hope, right? Like if AI like I know Matt's has talked about this

before at the you know Ruby Kayeggi conferences, you know, is like, you know, the Ruby R Ruby core team is gonna focus on okay, how can we make, you know, Ruby AI friendly and AI supporting tools, right? and I don't know. That's my take.

Ernesto Tagwerker (55:56.423)
Yeah, yeah, and I think I agree with that. I think there

Joe (56:02.926)
I think I think Ruby Central just kicked me out of of Riverside. Does Ru is this a Ruby at Riverside? I I finished my rant and then I was immediately ejected. So I'm back. Sorry about that.

Ernesto Tagwerker (56:06.569)
Ha ha ha.

Valentino Stoll (56:10.757)
Ha ha.

Ernesto Tagwerker (56:13.265)
Yeah, we lost you there for a second. I don't know. But yeah, what Valentina was saying that he he wants to see more from Ruby core related to AI and more or or better support and tooling for for Ruby, but also like more more people using Ruby to build like AI solutions.

Joe (56:15.118)
Ha ha ha.

Ernesto Tagwerker (56:35.0)
So I I agree with that and I I also understand like there's you know Ruby Central messed up with with RubyGems and and Bundler and all that and I think they're trying to like you know reform the the organization and like basically take it in in the right direction. But yeah, it's now it's like still like too soon to tell whether it is the right direction or not.

Joe (57:01.52)
What I'm sorry that I missed that, V. when I when I hear about, you know, supporting AI in open source, I think about you know, like our friend Carmine, you know, friend of the show, you know, Ruby LLM, I think about these you know, I think about things like NextRails, right? Like you know, tools that are either AI enabled or or AI focused that

work really well with with Ruby. Is that the kind of thing that you're talking about? That you'd like to see support for those kinds of things?

Valentino Stoll (57:35.258)
Yeah, I guess I'm torn, right? Like 'cause we have like this dichotomy of like Ruby tooling and ecosystem infrastructure, that has historically been like, you know, Ruby Central focused. And then it's like, okay, and conferences on top, right? and like what we're seeing over time is like conferences are like a big commitment, right? And they end up eating up a lot of resources, right? and so like, you know, do those should those two layers coexist, like I I feel like that's where a lot of the friction happens, right?

Joe (57:54.673)
my god, yeah.

Valentino Stoll (58:04.493)
And so I'm I'm wondering like, you know, Ruby and AI as like a, you know, concept together is really a marketing problem, right? And seems better suited for that layer of like conferencing and marketing and like, you know, l I was saying the Coca-Cola method of, you know, you see everybody drinking a Coca-Cola and you want to drink one is like the mentality of like where, you know, probing languages fit in the AI landscape, right? as far as what people reach for, what the AI reaches for, like those are all like

Joe (58:14.643)
I see.

Valentino Stoll (58:33.707)
you know, those marketing problems that so are have different solutions, right? Different paths, different people that would be better at solving those. so I wonder if it's a disconnect of that, right? More than anything. but I don't know. I'm I'm excited for conferences. So like I always get like great takeaways, meet great people. I I'm I don't know if I'm gonna be able to make

Joe (58:41.039)
Mm.

That's an interesting perspective.

Joe (58:52.814)
Me too.

Valentino Stoll (58:58.265)
This particular conference I would love to but I did make the last Ruby Rails conf which was fantastic in Philadelphia. And it was you know, it's just great to great to connect with people and like see what they're working on and like, you know, I I feel like you come away and build something new because of that, right?

Joe (59:05.839)
Yeah, it was great to see you there.

Joe (59:15.843)
Well yeah, and that's that you know what you you almost took the words right out of my mouth. I mean, for me, I just

It energizes me in a totally different way. And I you know, I go to the I go to the local meetups for the same reason, right? And they'll give they'll give me like a little charge every month. I go to I go to a big regional or big national conference and I get a you know, kind of a bigger charge. you know, and it sets me up for months after that. And I'm looking forward to this one for the same reason. And I'm actually I bought a ticket to Railsworld for the same reason. I haven't been to Railsworld and l you know, lo and behold, there's no more RailsConf, so Railsworld is here in the US this year, and I think that's great. You know, I'm I'm looking

Ernesto Tagwerker (59:29.689)
Mm-hmm.

Joe (59:51.705)
forward to it.

Ernesto Tagwerker (59:53.11)
Yeah. Yeah, I'll I'll be at Railsworld too and it's gonna be my first one as well. couldn't couldn't swing going to both of so I had to pick between RubyConf and Railsworld and I think I've never been to Austin, Texas, so I think that was one of my my main reasons to to go. but yeah, I love to to see if people, you know, want to talk about Ruby and AI at Railsworld, be happy to to talk about that.

Joe (59:58.447)
Mm-hmm.

Joe (01:00:21.027)
You know, and just a plug for for these conferences, and I know, you know, I think Ruby Central is not sold out yet, so you can still get tickets and I think we're I think we're publishing this before the conference. and you know

I don't want to toot our own horns too much here, but you know, the three people on this podcast are, you know, three very successful Rubyists that have been around for a long time and and a part of it is I see you guys at every conference. You know, Ernesto, you you've run Philly RB for I don't know, it feels like a hundred years and and yet, you know, you live out there in New Jersey, you still come over here, for Artificial Ruby. Valentino, you and I met at Artificial Ruby, you know, you're a speaker at conferences, you both are. You know, you go to conferences all the time, you support the community, but you do it because you get something out of it. And, you know, that's something.

is is not always easy to pin down, but it's easy to see in in retrospect that it it really has a a a large positive impact on on our careers.

Valentino Stoll (01:01:16.676)
Yeah, totally.

Ernesto Tagwerker (01:01:16.852)
Yeah. Yeah, and I think it's like a a great place to go and like I actually prefer to hang out on the hallway track. I can catch the talks later, I can watch them online. but I get the most energized when I talk to other people and I hear what they're doing in Ruby and Railsland and of course, yeah. I never miss your talks.

Joe (01:01:24.687)
Mm-hmm.

Yeah.

Joe (01:01:34.509)
Yeah, except when we're speaking. I know you come to those. Yeah, I know what you mean. I mean, 'cause yeah, these days I do I do a lot less coding and so, you know, can I go to the talk and learn a you know, learn about the next gem or about the next thing? I I could. and it's always interesting to me, but yeah, I I I love talking to people in between. I love seeing people I haven't seen in a year. It's great.

Valentino Stoll (01:02:01.049)
Yeah, totally.

Joe (01:02:02.785)
So we're coming up on time here, but it was it was really great to have you on the show, Ernesto. great to learn you know, about the evolution of Fast Ruby and about Ombu Labs as well. anything you wanna anything you wanna plug or or say before we sign off?

Ernesto Tagwerker (01:02:22.703)
yeah, I think it would be great if anybody wanted to give our clot skill a try. especially if you're running like a really old Rails version, would love to know if it worked for you. If it didn't, we don't have any sort of like telemetry code in there. so the only way I have to know whether people are using it if if you people are starring it on GitHub or telling me like at meetups, like you said. but yeah, I'd love people to give it a try. Open issues if they

they find any and hopefully it'll make their next upgrade project easier.

Joe (01:02:59.383)
Yeah. Thanks. And thanks and thanks also for the the great work you're doing with with open source. I mentioned NextRails, but you mentioned Ruby Critic a couple of times and I just wanna give a plus one for that. That has been a a fantastic gem for years and years. and you know, we're really happy that you steward it. We still use it. you know, to this day. It's it's a really excellent way of just getting to understand your code base and w you know, what you need to do to improve.

Ernesto Tagwerker (01:03:25.883)
Yeah. Yeah, it's a pretty cool evolution to metric foo. I remember the first Goruku I ever went to. I sat next to someone and they were talking about metric foo and I was like, that is so cool. I think it was a Pace University. yeah. And it was around the time I think when Rails two, three and three were merging, so anyway. But

Joe (01:03:38.159)
Is that in the basement at a Delphi?

Joe (01:03:42.77)
a pace, you're right, a pace, pace, pace, yeah. Yeah.

Joe (01:03:49.156)
Yeah.

Ernesto Tagwerker (01:03:52.323)
I I remember Metric Foo was such a cool thing and we actually are the maintainer maintainers of Metric Foo, but a lot of the tools that it used are kinda dead. So we might just focus on Ruby Critic from now on. but yeah, and if anybody loves Ruby Critic or wants to give it a try, it's like a great tool to learn where the complexity of an in a Rails application lives. So yeah.

Joe (01:04:03.033)
Yeah.

Ernesto Tagwerker (01:04:17.607)
Happy happy to maintain it and thanks for for your words. Joe. I appreciate it.

Valentino Stoll (01:04:23.087)
Yeah, plus one to that too. I I have used Ruby Critic for years as well. fantastic. Yeah.

Joe (01:04:25.485)
Yeah.

Ernesto Tagwerker (01:04:29.862)
Cool. and a small announcement. Now JRuby is officially supported by Ruby Critic. I made sure of that last week. So yeah, Charles Snodder and and I have been talking consistently and he's helped me be a better JRuby citizen. So quick announcement there for you.

Joe (01:04:37.158)
excellent.

Valentino Stoll (01:04:38.106)
that's awesome.

Joe (01:04:48.835)
Good. Good. That's great.

Valentino Stoll (01:04:50.915)
JRuby is fantastic. I've used that before on projects and yeah. if you have a Java like world organization that you're supporting, you can use Ruby. Yeah. Yeah.

Joe (01:04:54.221)
Mm-hmm. Yeah, me too.

Joe (01:05:00.931)
Yeah, stop writing Java. There, JRuby exists. Yeah, the well grounded Rubyist. It's long awaited publication date is imminent. So probably not maybe by the time this goes live, but probably about a week later. so check it out.

Ernesto Tagwerker (01:05:04.315)
Yeah.

Ernesto Tagwerker (01:05:20.869)
That's great. And thanks for all your work on on that as well. it's one of my favorite Ruby books, so it's awesome to to be in touch.

Joe (01:05:26.391)
thanks. that is a labor of love and it was great to be able to you know, hook up with David Black again and and work with him on this. Is I mean he's he's just brilliant and his writing is incredible. So

Valentino Stoll (01:05:37.187)
That's awesome. Yeah, I've got one one small, small announcement. it's just kind of fun. my bot Minerva has has created her own website, Minerva.codename V dot com, to journal our progress and reinforcement learn off of that. so if you are interested in what my bot is doing for me, you can go there and find out. it's pretty funny. But

Joe (01:05:39.811)
All right, everyone. yeah, go ahead, sorry. Yeah.

Joe (01:06:02.85)
awesome.

This is your wa this is your agent blogging? I love this.

Valentino Stoll (01:06:06.881)
anyway.

Yep. Yep. So kind of just a fun project. It you know, it she brought it up and she it brought it up and and I thought, that's a great idea, you know? Yeah. All right, well th thank you so much for for coming on, Ernesto. This has been fantastic. And hopefully we get to meet in person and we we could talk about all kinds of stuff.

Joe (01:06:18.305)
Hey, you know, yeah. Yeah, give give the lady what she wants.

Agent that can't wait. Yeah, thanks Ernesto.

Ernesto Tagwerker (01:06:34.412)
Yeah, yeah. Thanks for having me. It's been a pleasure. have a good one.

Valentino Stoll (01:06:37.434)
Yep. All right. Thanks.

Joe (01:06:39.203)
All right, guys. Thanks.

### Modernizing a high-stakes system for AI-assisted development?

The AI Modernization GUIDE Assessment helps teams understand where AI can safely accelerate engineering, where modernization is needed first, and what evidence should guide the roadmap.

[Start with a GUIDE Assessment](https://www.defmethod.com/assessment)

[← Back to all episodes](https://www.defmethod.com/podcast)