I gave Claude Code our entire codebase. Our customers noticed. | Al Chen (Galileo)

Al Chen is a field engineer at Galileo, an observability platform for AI applications, where he works on the front lines with enterprise customers asking highly technical questions. Despite never having held an engineering role, Al has built a system using Claude Code to query Galileo’s 15 separate repositories, combine that with Confluence documentation and customer-specific quirks, and deliver hyper-personalized answers that would otherwise require constant engineering support. What you’ll learn: How to use Claude Code to query multiple repositories simultaneously for customer supportWhy code is often a better source of truth than documentationHow to combine repository context with Confluence and Slack using MCPsThe “customer quirks” system that creates hyper-personalized deployment guidesHow to build virtuous loops that turn single customer questions into scalable knowledgeWhy information organization matters less in the AI eraA simple 16-line script (written by Claude Code) that pulls the latest main branch across all your repositories to keep your context currentHow to reduce engineering interruptions to near-zero by empowering customer-facing teams to query the codebase directly— Brought to you by: Orkes https://www.orkes.io/—The enterprise platform for reliable applications and agentic workflows Tines https://tines.com/howiai—Start building intelligent workflows today — In this episode, we cover: (00:00) Introduction to Al Chen (02:50) The problem: documentation wasn’t enough (04:23) Pulling 15 repos into VS Code (06:03) How Claude Code queries the entire codebase (08:00) Why current code beats documentation (08:31) The pull script that keeps everything updated (09:54) Opening projects at the multi-repo level (11:40) Live demo: answering deployment questions (13:25) The customer quirks system (15:00) Living in chaos: why organization matters less now (17:03) Competing on customer experience, not just product (18:20) Should customers be able to query the code directly? (20:05) Where humans still add value (25:46) Using AI for reactive Slack support (29:16) The “and then” workflow discovery (32:07) Scaling processes across the team (34:07) Lightning round and final thoughts — Tools referenced: • Claude Code: https://claude.ai/code • VS Code: https://code.visualstudio.com/ • Pylon: https://usepylon.com/ • Confluence: https://www.atlassian.com/software/confluence — Other references: • Slack: https://slack.com/ • Kubernetes: https://kubernetes.io/ • Stack Overflow: https://stackoverflow.com/ • Intercom: https://www.intercom.com/ — Where to find Al Chen: LinkedIn: https://www.linkedin.com/in/thealchen/ Company: https://www.rungalileo.io https://www.rungalileo.io/ — Where to find Claire Vo: ChatPRD: https://www.chatprd.ai/ Website: https://clairevo.com/ LinkedIn: https://www.linkedin.com/in/clairevo/ X: https://x.com/clairevo — Production and marketing by https://penname.co/. For inquiries about sponsoring the podcast, email [email protected].

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  The minute I realized I couldn't really do my job was when I was trying to reference our public documentation and trying to provide an answer, it just still wasn't coming up with the answer that my customers were looking for.

• 0:11Claire Vo?

  They don't want the docs answer. They want the step-by-step answer of how all these services cascade together.

• 0:17Speaker 00

  What I realized is that I can actually pull all of these repos into my VS code, and I can now use Claude Code to ask our entire codebase questions.

• 0:26Claire Vo?

  Did you just say Claude Code? Write me a script that pulls all these.

• 0:30Speaker 00

  Yeah, I'm opening up the script right now. It's like, what, 16 lines? Didn't have to write this. I just said, "Help me figure out a way to pull the latest main branches into my local repos."

• 0:38Claire Vo?

  The reality is, we can now all live in a little bit more chaos because the AI navigates all that information for us across systems, right? So you can be in your code, querying Confluence, we'll find the information. You have to be less precious about where and how you store the information.

• 0:55Speaker 00

  Throw it into Confluence, throw it into Notion, throw it into Slack, whatever. That ends up being context you can provide to Claude when you are trying to ask it a question about a customer or about your codebase.

• 1:06Claire Vo?

  Let's give Claude Code a little spiff every time it answers a question correctly. You gotta split your quota with Claude Code.

• 1:13Speaker 00

  Yeah, it gives you better answers, the more bucks you give it or something.

• 1:16Claire Vo?

  Coin-operated Claude, that's gonna be my new skill.

• 1:22Claire Vo?

  Welcome back to How I AI. I'm Clara Vow, product leader and AI obsessive, here on a mission to help you build better with these new tools. Today, we have an episode all about harnessing your code to make your customers' experience way better. Al Chen, who's on the field engineering team at Galileo, shows us how he uses their 15 repositories and Claude Code to answer every nuanced customer question that comes across his desk and use that to make the entire customer base and his entire team a lot happier. Let's get to it. This episode is brought to you by Orkus, the company behind Open Source Conductor, which powers complex workflows and process orchestration for modern enterprise apps in agentic workflows.

• 2:05Claire Vo?

  Legacy business process automation tools are breaking down. Siloed low-code platforms, outdated process management systems, and disconnected API management tools weren't built for today's AI-powered world. Orkuss changes that. With Orkuss Conductor, you get a modern orchestration layer that scales with high reliability and brings humans, AI, and systems together in real time. It's not just about tasks, it's about orchestrating everything. APIs, microservices, data pipelines, human in the loop actions, and even autonomous agents. So build, test, and debug complex workflows with ease, all while maintaining enterprise-grade security, compliance, and observability. Orkuss. Orchestrate the future of work.

• 2:47Claire Vo?

  Learn more and start building at orkes.io. Al, thanks for joining How I AI. I am really excited about this episode because we've seen a lot about using your code as documentation. You know, we've heard engineers saying, you know, docs and code should be in the repo, product managers saying code can now be my documentation for internally facing assets or as I help draft PRDs. But you're going to show us how you can use code as an asset to create customer-facing things and solve customer-facing problems. So tell me, what problem were you facing when you decided,

• 3:28Claire Vo?

  "I'm just going to clone the repo and fire up Claude code and solve some of these problems myself"?

• 3:33Speaker 00

  Sure. So working at Galileo on the field engineering team, I'm on the front lines in terms of working with our enterprise customers, who are typically developers themselves and asking very in-depth technical questions. And the minute I realized I couldn't really do my job was when I was trying to reference our public documentation and trying to provide an answer to my customers, even by using Claude code or ChatGPT or whatever, and trying to take all these different help docs and trying to come up with the answer, it just still wasn't coming up with the answer that my customers were looking for. And I just background, I'm not an engineer.

• 4:16Speaker 00

  I've never held an engineering role, but I think I know enough to just be dangerous. And I realized that our product, Galileo's product, we're an observability tool for AI applications. If you look at this image here, I'm showing an architecture diagram, high level of all the different services that make up our platform. This is all like backend images that you have to, that customers have to deploy onto their Kubernetes cluster. And I realized that all these different services like UI, API, AuthZ comment, they are all individual repos within our Galileo repo. We're not a mono repo. We have multiple different repos.

• 4:58Speaker 00

  And so what I realized is that I can actually pull all of these repos into my VS code. Initially, it was just more for me to like, I want to understand like how our code works and how our code is structured. But then when I threw it all into VS code, which looks like here, you notice along the left-hand side, I'm open VS code now, and most of these directories correspond to one of those services within our architecture. So one repo corresponds to one service. And by having all of these repos in my VS code, I can now use Claude code to ask my, our entire code base, questions that are not answerable by our public documentation.

• 5:43Speaker 00

  And so sometimes I'll get really end up questions about, oh, how does this feature actually work? And so I'll ask Claude code, look into the API repo, look into the AuthZ repo, and help me come up with an answer. If you can't find the answer, reference other repos within my directory, my core, my root directory, and help me figure out the answer. And so that's the key unlock was when I figured out I could get way more in-depth, way more technical, and at the same time, myself, I can learn how our codebase works. And how this has really helped me is I don't have to constantly ping our team engineering channel with, hey, what's the answer to this question? The customer just pinged me about this. And you can imagine engineers being really frustrated when I'm trying to post

• 6:28Speaker 00

  these questions and then the customer asks me a follow-up, then I'm posting that follow-up in the Slack thread. So I'm sure many of you who are working on the front lines of customers understand how that feels. But I've basically reduced all of that almost down to zero by pulling all these repos into my local VS code.

• 6:46Claire Vo?

  I'm really empathetic to this problem because I used to work at Launch Darkly, leading product and engineering, very technical product. We too had an architecture diagram that looked very similar to that. And again, as a more of a, you know, people think that CPOs, Chief Product Officers, or CTOs are internal facing. No, no, no, we're salespeople. We're always salespeople. You trot us out and you put us in front of the customer or you put us in front of the prospect to answer the technical questions. And we had a diagram like that and I would constantly get these very detailed questions that required very detailed answers. Like, how does your caching work? And you know, when you have seven layers of caching in your app, you can give the high-level docs answer, but when you're sitting

• 7:30Claire Vo?

  with, you know, an architect in the room or somebody highly technical, they don't want the docs answer. They want the step-by-step answer of how all these services cascade together to build a resilient caching mechanism, for example. And I just think how powerful is it to be in a meeting or in an email back and forth and not just sort of give this high level, but be able to query the current codebase and really understand at a detailed level how it works. And I think current is very important because you know and I know, this is always evolving over time. So even if you got the answer right, you know, a month ago, maybe your team shipped an update, or maybe, you know, that method is actually

• 8:14Claire Vo?

  out of date, or the docs are a little bit out of date. And so I do think because the codebase is, you know, at least your main branch is always the source of truth, it becomes a really reliable, you know, context set for you to answer questions about how the product operates.

• 8:32Speaker 00

  Yeah, and to quickly address your comment about how your code is obviously always evolving. I mean, we're pushing out, you know, multiple features per day, multiple releases. And one thing I've done, wrote this with Claude Code, is I have this script at my root directory that says like, I just do something called pull all. And I'm not sure if this is how other people do it, but it just pulls the main branch into my repo for all the repos in my root directory. So if I do this every day, I kind of get the latest code across like all these directories on the left-hand side of my VS Code. So the alternative, which I was doing before for like a few weeks, and I realized this is just asinine, I'm doing this, is doing git pull origin main on every single directory.

• 9:17Speaker 00

  And it was just like not scalable because there's now like 15 different repos I have to pull the latest from. So that's kind of how I solved the code base is always evolving problem to make sure that I'm always getting the most up-to-date information for my customers.

• 9:32Claire Vo?

  And I have to ask, did you just say Claude Code, write me a script that git pulls all these?

• 9:37Speaker 00

  Oh, yeah, yeah, yeah. I have no idea. This is, I'm opening up the script right now in my VS Code, and it's like, what, 16 lines? I didn't have to write this. I just said, help me figure out a way to pull the latest main branches to my local repos. And it just did it in like one shot.

• 9:54Claire Vo?

  Yeah, the other thing I want to call out for folks as I'm looking at your screen is I don't think people use this trick enough, which is in VS Code, in Cursor, and whatever your IDE is, loading a project at the multi-repo level as opposed to at the individual repo level, if you're trying to answer questions across the product, is really important. So, you know, there's some like context bloat stuff that can come into sort of querying across all those repos and all those files. It would be very painful if you had to go into each of these repos one by one. And like query and then go into the other one and query. And so I like this idea of opening

• 10:37Claire Vo?

  them all jointly in in your IDE so that when you're querying it with Claude code or you're clearing it, you know, with something like cursor, it can have, it can, it can go across, traverse across repos and really give you highly contextualized answers.

• 10:54Speaker 00

  Yeah, our code basis happens to be in multiple repos, but I just pulled them all into this like giant Galileo directory here. And so everything is like at the same parent. But yeah, if you're in a monorepo, could be, yeah, I actually, I don't know how this would work with a monorepo because I've never done it with monorepo with Claude code. But at least for us at Galileo, this is how how it works.

• 11:15Claire Vo?

  Well, I have a many monorepos and yeah, you just open it at the right. I would, I would say my advice to folks is open Claude code or open, you know, your IDE at the right level. And sometimes it's narrow and sometimes you need to go up a directory. And I think really thinking about that and you can even do that contextualize to the problem you're trying to solve. Right. And and doing that, I think, is really helpful. Can you show us just using Claude code, what kind of question you could answer with this code context?

• 11:47Speaker 00

  I will give you an example of, I guess, I'm a big believer in using shortcuts too. So I use a bunch of custom Claude code custom commands to help me do stuff. So one thing I do a lot is helping my customers deploy Galileo into their VPC. So I have a custom command called DPL, which is, it actually references our, the first thing it does is it looks at our Confluence because we have a whole bunch of Confluence pages about how to deploy into Kubernetes using our different images and stuff like that. So I'll say DPL, my customer cannot use CRDs and they are using

• 12:31Speaker 00

  Google Secrets Manager and want to deploy the wizard image.

• 12:42Speaker 00

  Give me a step-by-step process on how to do it. This is actually not a super representative query because they're way more detailed than this. I provide a lot more context. But I want Claude Code to focus on looking at Confluence first because I know that we have a whole bunch of deployment stuff there. And then from there, if they can't find the answer, it will go off into all of different repos along the left-hand side of my Claude Code VS Code to find the answer. So right now, it's just using the last C and MCP to pull information from Confluence and then marries that with like our code base to answer a very kind of in-depth deployment question.

• 13:24Speaker 00

  The one thing, I'm not sure if we should talk about this now, but I started doing this in Confluence where we have a, we call it a customer quirks page. These are all kinds of, all of our enterprise customers, you typically have air gapped environments, so they have all these security measures and we have to abide by them when we deploy the product into their environment. And so I literally have a page that looks exactly like this where I have the customer's name at the top level and then a bunch of bullet points with like, you know, here are some things about how they store their secrets. Here's how they do namespaces, you know, here's how they handle sidecars and services, service-to-service encryption, things I have no, I know nothing about.

• 14:08Speaker 00

  But as I'm meeting with my customers, I'm putting this all into this one Confluence page, this ever-growing Confluence page. And then this is actually one of the core pages that goes into this DPL custom command, which is look at the customer quirks page. If I'm mentioning a customer that's on that page, look at all their quirks. And then in the response from Claude, it's highly customized, highly tailored to their environment because I've seen from working with our DevOps team that We can provide a generic answer about Kubernetes, or about ClickHouse, or about whatever for the customer, but it's like something you can just find online by Googling or using AI.

• 14:50Speaker 00

  But when it's tailored to specific security requirements and deployment requirements, it's way more effective. And this gives the customer more trust that we know what we're doing, essentially.

• 15:01Claire Vo?

  What I love about what you showed here, which is, you know, kind of combining the repository with the Confluence MCP, and then both like team-generated general documentation, as well as you-generated like micro-documentation at the customer level, is I've heard so often in my 20 years in enterprise SaaS, like, what is the source of truth for this information? Like, I'm sure you've heard this too. Like, what's the source of truth for how XYZ works? Or what's the source of truth for this customer? And people have spent so much time, like, you know, pruning these Confluence gardens and organizing their Slack channels and trying to get people to, you know, get naming conventions, right? And like, the reality is, we can now all live

• 15:43Claire Vo?

  in a little bit more chaos, because the AI navigates all that information for us across systems, right? So you can be in your code, querying Confluence, it will find, you can kind of point it in the right direction, it will find the information. You have to be less precious about where and how you store the information, bullet point list of quirks, you know, like, really official docs, whatever, it doesn't matter, because AI is just so much more effective at traversing all that information and pulling it in and making it actionable for you. And I don't think that's anything like any human was really proud that they were good at. They're like, I'm really good at finding like the right Confluence doc. That was never, never the value add.

• 16:25Speaker 00

  Yeah, yeah. I mean, I think even if it's as simple as, hey, you came across a really great answer in Slack, like in a really engaging Slack thread, throw that into a Confluence page, or save that Slack thread, because I also use the Slack MCP to be able to summarize threads. So if you have like just some random, like this ongoing stream of cautiousness of documents you want to have Claude Code scan, I would just say, throw it into Confluence, throw it into Notion, throw it into Slack, wherever. And then that ends up being context you can provide to Claude when you are trying to ask it a question about a customer or about your code days.

• 17:04Claire Vo?

  Well, and the other thing, and this is maybe going back to how I introduced this episode, which is people use AI so much to compete on the field of the product and engineering velocity. And what I mean by that is like, we're all using Claude code to ship more product. We're all using AI and Codex to build, you know, better user experiences or more resilient backends or any of that stuff. But there's also a completely different competitive field, which is how you show up in your relationships with your customers. And I think, you know, what you're showing is you can actually use AI to invest and compete on customer experience. And, you know, my hypothesis is when your very complex enterprise customers have you

• 17:46Claire Vo?

  show up and you don't just say like, here are our general docs to deploy this. And instead you say, I heard you, I understand what your needs are, and here are your custom docs on how you specifically need to deploy this. And I've already pre-thought about all the problems you've already told me about. You know, just looking like in a competitive sense, that's got to come across as a much more enjoyable customer experience on the receiving end and allows you to position yourself, not just as a great product, but as a great team that's going to service your customers well.

• 18:22Speaker 00

  Yeah, I hope so. I mean, I think our customers, I think my customers are hopefully enjoying the answers I provide and the in-depthness that I provide. I think I've thought about taking this to the extreme, which is that we have certain, I've certain customers who are like, you know, very in the weeds. They want to know things like right this very second. And I'm literally taking their question and then just like saying, my customer then asked me this because they can't see your code, but me, Al, I can see the code. Help me get the answer. And so if I take that to the logical, logical conclusion, it's like, why can't we just share our repos with the customer? Because then they can just start querying our repos directly to get the answers they need instead of me as kind of like the quote

• 19:04Speaker 00

  unquote middleman. And, you know, the issue is that like our code is proprietary and all that kind of stuff. But I have seen, there's actually a case study from Langchain. And since a lot of Langchains Um, repos are, you know, it's open source.

• 19:19Claire Vo?

  Uh huh.

• 19:19Speaker 00

  Like their support agent bot actually does a lot of things I do, but it is able to query, you know, all the public open source repos. And any of you out there who are trying to use Langchain or Langgraph, you can just pull all those repos down to your local machine and then ask questions, of course, using Claude Code or Cursor or whatever. Um, but I've come through that kind of thought experiment of like, I'm still kind of a bottleneck in terms of answering my customers' questions because I kind of like hold the keys to our code. But if they somehow had a sanitized version of it, then maybe they could just self-answer their questions too, because they're also all using VS Code and Cursor and Claude too,

• 20:02Speaker 00

  but they just don't happen to have our, you know, proprietary code base.

• 20:06Claire Vo?

  Yeah, I was, I was gonna ask you, are you worried that like the Albot is coming and you're, you're cut, you're cut out of it? And I, I'm just curious how you think about then? When like, you, you, again, like the highest order of you is not to be a pass-through, and I don't think you think of that yourself as that. And so where does the human in these relationships powered by AI, you know, add, add the value?

• 20:30Speaker 00

  Well, to, I don't just blindly copy and paste the answers I get from Claude Code to my customers in Slack or email or wherever. I still try to proofread everything and I actually do like try to make it sound more human and you can then say the argument, oh, why don't you use Claude Code to make your answer sound more human? And I think all of us know when we get an answer that's from AI and it's, you know, things like, you know, you'll see like a bullet point saying like, in summary, here are the things you need to do to make sure your ClickHouse works within. So it's like removing things like that that just make it seem like it's from a bot just makes it seem more human. And we've actually, I mean, this is kind of going, going behind the scenes of how we work,

• 21:14Speaker 00

  but we've been dinged sometimes where the customer will say, can you just not give me an error response and just give me like a human proofread of it and tell me how it applies to me? Because typically the response is way too verbose, it has way too much information, and the customer just wants to know, give me like the bottom line upfront, what do I need to know to like deploy this image onto my cluster? And so. That's where the human, I still see myself as a human providing value and calling that down to what they actually need. And I would say even for some of the more in-depth technical questions, I still try to get an engineer's perspective on it to make sure like Claude Code is not hallucinating

• 21:57Speaker 00

  or not saying anything out of the ordinary. In my system prompt, I always, you know, in my Claude Code, I say things like, don't make anything up, always cite your sources, point me to the line of code where you're getting this information from. But even with that, if I don't fully understand how this function works, whatever, I'm still paying like the engineering channel to say, hey, this is what Claude Code told me, does that jive with what you're thinking? And there are times when I'm wrong or Claude Code is wrong because engineers have been thinking about refactoring into this new model, which is not captured in our code base anywhere. It's just captured in like a meeting note somewhere or just like, you know, hallway conversations.

• 22:39Speaker 00

  And so those are the things that I'll never be able to query, let's say, in Claude.

• 22:44Claire Vo?

  Yeah, I would, I would say the other thing that, you know, where I see humans adding value, and I say this all the time, which is like RIS is the only moat, which is at some point, you know, people just want to have a face and a trusted personal relationship, you know, with the folks, and this is like my enterprise showing, but like with the folks that are selling them software, you want to know that you have somebody to call. You want to know that you have somebody that can gather the right, the right folks around your team and your deployment. And, you know, you want to enjoy working with that person. And I will just say, I get a lot of, it is very fun for me to build with these tools,

• 23:24Claire Vo?

  with AI tools, but I wouldn't say my AI colleagues are like the most fun to hang out with, which is like, I'm not like always looking forward to like my, my Claude Code session, like I'm going to really chit chat with, with good old Claude. And I do think you still have that relationship with. You know, your human partners, your human colleagues, all that sort of stuff. And so I think there is a piece of that that's just not going to get cut out. And honestly, I gave this talk, I don't know, two years ago, I said PM is dead. And people are like, well, what else should we do? And I was like, get into sales. Like, that's not going away. Customer facing stuff is not, is not going away. So for anybody that wants to survive, you know, the, the incoming apocalypse, I do think customer facing roles and spending more time customer facing is a really important part of everyone's job.

• 24:10Speaker 00

  Absolutely. If you're working enterprise sales, like, that is all people, handshakes, lunches, dinners. So that will never be replaced, I think, by AI anytime soon.

• 24:22Claire Vo?

  Well, you know, and there might be a generational shift, though, here. I think as, as we sell, as we sell, we'll see. You know, I used to say my, my joke in enterprise sales and the biggest, the, the biggest headwind to enterprise sales was I was starting to sell to millennials who like wanted you to text when you showed up at their door. They didn't want you to knock on their door like they're

• 24:46Claire Vo?

  we'll see, we'll see how enterprise sales changes generationally.

• 24:49Speaker 02

  This episode is brought to you by Tines, the intelligent workflow platform powering the world's most important work. Business moves faster than the systems meant to support it. Teams are stuck with repetitive tasks, scattered tools, and hard to reach data. AI has huge promise, but struggles when everything underneath is fragmented. Tines fixes that. It unifies your tools, data, and processes in one secure, flexible platform, blending a jet-tick AI, automation, and human-led intervention. Teams get their time back, workflows run smarter, and AI actually delivers real value. Customers now automate over 1.5 billion actions every week.

• 25:33Speaker 02

  Tines is trusted by companies like Canva, Coinbase, Databricks, GitLab, Mars, and Reddit. Try Tines at tines.com/HowIAI.

• 25:47Claire Vo?

  All right, so we have, just to recap, we've shown how you use all these repositories in your very complex codebase. Pair that with Claude Code, which is made more efficient through a couple like shortcuts and scripts to be able for you to answer customer queries and then also build custom deployment plans for your customers anchored in exactly how your code works and exactly how their infrastructure works, making everybody happier and getting customers off the ground quicker. But there are also instances where you need to be doing more reactive support in different channels. And I know you're using AI for that. So you want to walk us through

• 26:29Claire Vo?

  how you're using AI and Slack and supporting customers there?

• 26:33Speaker 00

  Yeah, so like many, I say, digital AI native companies, we do a lot of our customer support through Slack. You know, we have external channels with our customers and not, I mean, I come from the, I used to work in a world where everything was through like a central Zendesk or Intercom or whatever. But for enterprise customers, it's kind of like a on the go, always kind of on kind of thing. And so we use a tool internally called Pylon for monitoring all our different external Slack channels. And I'm going to show you what this looks like in this tab. And this is an example of a conversation I had with a customer asking in-depth questions about like our Galileo callback function and how it

• 27:16Speaker 00

  emits different events. And as you can imagine, I was using Claude Code to help answer these questions in addition to using our docs. But when you're looking at a conversation like this in Pylon or in Slack, the first thing you have to think about is like, I wonder if like, I could turn this into a help article, or if I should update our docs, or will other customers benefit from the knowledge that's being trapped in this little Slack thread? And so what Pylon allows us to do is looking at a really long Slack thread. It can help you generate a help article. And right here, I already have one that's associated with this specific conversation. But it's literally just clicking on add article, generate article draft, and then we have these different templates, and it just creates

• 28:01Speaker 00

  like this article for you on the fly. Now, this is not rocket science. You could copy and paste the whole Slack thread, put it into Any AI tool you want to generate help article. The main thing with pylons is everything is kind of just like in one interface. So you don't have to worry about like copying and pasting and putting links together. So this is kind of like that draft that this came up with. And then we have this ongoing list of articles based on real customer conversations. And those articles are abstracted to not show any specific customer information. But then when we publish these articles, they go into this knowledge base, which is also public knowledge base. And this is kind of like the living truth of like in-depth, in the weeds.

• 28:46Speaker 00

  Questions about deployment, about how Galileo works. And it's always way more in-depth and way more up to date compared to our docs. Because our official docs require pulling down the docs repo, submitting a PR, getting it approved, so on and so forth. And so it's a lot more of a polished process. Whereas with these knowledge-based articles, it's kind of like just on the fly. You have a Slack thread, you want to summarize, use it, create it in Pylon. And then it just automatically gets auto-published to this knowledge-based site.

• 29:17Claire Vo?

  So one of the things that I love about this is this represents my, like what I call the and then workflow discovery in AI, which is, I say, imagine you had an infinitely staffed team. And you were faced with a task and every time they did one step of the task, you asked and then, and they were able to do it. So it's like, I got a Slack query from a customer. So I answered it and it was like, if you had a perfectly staffed team, what would you do next? And be like, and then I would turn that into an article. And it was like, okay, and you turn it into article. And then what you do is like, and then I would share that with our customer success team and train them on this answer. Because, you know, everybody needs to know this information and be like, and then, and you'd be like,

• 30:01Claire Vo?

  and then we could probably do like long tail SEO off all these questions. And I think you can like chain, chain these like and then, you know, workflows to actually build out like a pretty cool, you know, virtuous cycle system based off a single action. And because again, like the cost of doing any one of those collapses to zero. You can really pull the thread of these tasks that like no human team would have the capacity to really do. But if you think of it as a system, it helps your human teammates, it helps your customers, and you can get a lot of stuff done. We have a couple of episodes. Matt at Suzie showed kind of a version of this where he takes a customer,

• 30:43Claire Vo?

  a recorded customer call and like is bidding on AdWords for like phrases the customer says and like spinning up blog posts and doing like sales coaching off of it. So I think this is like a very similar example, which is you have this like, you know, atomic unit of a question in Slack and you've turned it into something that benefits, benefits the full team.

• 31:04Speaker 00

  Yeah, I think if you go back to pre-AI days, and I'm redoing this with Entercom, was you, we wanted to see whatever, what are users talking about the most when they're asking us questions? And so if you start clustering all these user questions and insights into different themes and categories, those can end up determining, determining your product roadmap too. And so I think with AI just kind of automates a little bit more of that without you having to like do the manual sorting, grouping within like Google Sheets or whatever. I know there's like platforms you can buy that do this for you. I think there's one called Interpret, which I've used in the past.

• 31:43Claire Vo?

  They've been a HowAI sponsor, so thank you, Interpret.

• 31:47Speaker 00

  Yeah, so, but, you know, I think, again, depending on how you want to view this whole virtuous lifecycle, maybe you don't want all of your data to be like in a silo in one place and you want it to be more open. So there's that to think about too. But yeah, AI definitely helps, to your point, make that virtuous cycle for customers, but also for your product.

• 32:09Claire Vo?

  So I have a question. Is this the the owl system? Or is this the Galileo, you know, field engineering system, which is, you know, you have this great workflow, you've discovered all these things. How does this sort of process get scaled out, shared, taught throughout the organization so that everybody that interacts with customers is benefiting from all the tips and tricks that you're figuring out yourself?

• 32:32Speaker 00

  Sure. So I, my previous background was I've worked in kind of the no-code, low-code space. And I'm a big believer in systems, tools, processes, and the tools that help you create those things. And so when it comes to, is this the Al way of doing things? Yes, it's my way. But I'm also very, probably one of the more opinionated people on the field engineering team about like, how we should be doing things in terms of talking to customers, answering their questions, and pulling in the right context. And so I've told multiple people, like, pull all the repos into your local machine, and have Claude code run an init command to index the whole code base, or whatever. And I'm

• 33:13Speaker 00

  just like constantly sharing these tips and tricks to my teammates to make sure they're also functioning at their capacity. So it's my way, but I would say I'm also very opinionated about how we should do things, because I've done things the hard way, the manual way. And this way to me is like, just 10 times way more productive. So we don't have like a specific, like, oh, because Al's doing it, but the whole team has to do it. It's more just like, people show, here's the problem I had, here's the results I had with Claude code, or whatever, and here's why I think you should adopt my solution. And I'm constantly having that conversation internally about like, how do we break out of certain processes

• 33:54Speaker 00

  that I think are slowing us down, and how AI can be infused into all those processes as well.

• 34:00Claire Vo?

  Well, and now you're sharing to all of our How I AI audience on how they can do that. So you're having more impact than just on your team. All right. Well, so to just recap again, your code is your source of truth. It can help you answer customer questions. It can help you document customer solutions. You can also do that with other channels like Slack, and then like, create these virtuous loops of solving a single customer's problem, and then a system to solve that problem more scalably across your entire customer base for yourself and for your teammates. Very, very high-impact episode. I think people are gonna have a lot of takeaways from this one. Super practical for all my friends that are customer-facing out there on things they can do, starting tomorrow, to use AI to

• 34:45Claire Vo?

  give their customers a better experience. Let's jump into lightning round questions and I have one that's really top of mind, which is, it seems like you have a very healthy culture at Galileo, but I can imagine teams, especially engineering teams, that are like, oh, no, no, no, no. I don't really want the customer-facing folks like going into our repo, querying it, and then just YOLO-ing answers over to our customer base, especially in a more technical product that really requires deep technical understanding. I think you've proven that there's a lot of value in doing that, but what would you say to those teams that are a little bit more hesitant about ungating access to the repo to non-technical roles?

• 35:28Speaker 00

  I think from the engineering engineer's perspective, I would look at it as how many, I would try to think about how many times in the last week, in the last day, have you been asked a last-minute question on Slack, a last-minute DM ping, mentioned in a thread where, how does this thing work? Where is this, how do we make sure that this is functioning the way it should be? And you're constantly the source of, you're the bottleneck for answering that question. And if you provide a system, kind of similar to what I have, to your, your customer-facing team, then you kind of just take away that toil and the constant like uncalmness of like answering these random product and engineering questions

• 36:13Speaker 00

  that is already in your codebase, or maybe it's already living in your confluence or something like that. So I think that's really the biggest takeaway for me is, how much of your time is being sucked away from your customer's team because they don't have access to the code? And, I mean, I think there's some no code, I mean, I think you can maybe pull in your code into Claude Co-Work, which is a little more no-codey and other kind of like more no-codey ways of doing things, but I think what I've shown is, I think, the most performant way of being able to pull your code and get answers out. So I think that's kind of a... From an engineer's perspective, how much time can you save?

• 36:53Speaker 00

  And then also how more effective your your customer facing org can be? And I think the corollary corollary to that is that our field engineering team is very technical. And so maybe you increase the hiring bar for your customer success or customer engineering team to feel comfortable using GitHub and pulling repos into your local machine. And so that could be today, if it's if they're not technical, just doing like a simple tutorial or enablement session on how do you use GitHub? How do you use Git commands? Things like that. And there might be some self learning you have to do on the side too. But I think once you're, once you have your environment set up, that's always like the hardest

• 37:37Speaker 00

  part about this whole exercise, getting your environment set up. Once that thing is set up, then using Claude code is just like using any other AI chatbot. So I think there's a few different ways I'd approach it from to democratize access to your repos.

• 37:52Claire Vo?

  One of the things I was going to say is I often tell people this is the era of the hard skill, which is no matter what role you're in, sorry, babe, you gotta like learn a little bit how to code. You have to learn a little bit what Git works like. You have to be okay opening up some code you don't understand in an IDE. Because that's just going to be the substrate by which we communicate for the next three years. It's going to go like closer and closer to the code. Because these LLMs are extremely good at understanding code. And so I think across the board, people just need to become more technical and develop hard skills around code, even if your job is not code.

• 38:31Claire Vo?

  I think the second thing that I tell people is there's no better time to learn how to code. Truly no better time to learn how to actually code. And I think people that are shipping with Claude code, but not using that as an excuse or a support to learn some fundamental software engineering concepts are missing 50% of the value. Like I had to teach myself how to code out of a book, like literally out of a book. It was, I had a book open and then I would look at my single screen because none of us had two screens, that would be crazy. And I would like read the book and I would type the book in the word, like the words in the book in code and press enter and it would say hello world. And that was my life.

• 39:16Claire Vo?

  And now you have this like magic. Super patient, infinitely wise, you know, like teacher in your computer that you can use to learn to code. And I think, you know, you talked a little bit about Kubernetes and how you you scaled up on that. So I'm curious your thoughts on just up leveling technical skills using some of these these tools.

• 39:38Speaker 00

  I think the meta takeaway is like, you just have to be curious about like how things work. I can't really say anything else besides that. It's kind of like, I come from that same world too of like looking at a book. And then I would say the graduation above that was knowing how to write a good Google query.

• 39:55Claire Vo?

  Yeah, Stack Overflow.

• 39:56Speaker 00

  Going to Stack Overflow. And then how many of you are listening with this where you go to some Stack Overflow Q and A, you copy and paste the code. It doesn't work perfectly, of course. So you're Googling the error you get from that thing you copied and pasted. And of course, everyone in Stack Overflow is super snarky and it's like, not a kind of healthy conversation. And then to your point, you have this like infinitely patient, infinitely kind assistant who never gives you the wrong code snippet from Stack Overflow. It's always like tailored to back to everything I'm saying, it's tailored to your needs to what you want. And then if you go the extra step, and like what you said, and then what?

• 40:38Speaker 00

  If you go the extra step and say, okay, thanks Claude, you told me this is the answer. Tell me why this works. And then it's you start getting into Kubernetes and into the deeper in the in the weeds things. But of course, you're not going to know everything rough the bat. So you can say things like, oh, explain to me in simple terms, explain to me like I'm five. And so you're just kind of pulling on that thread. And I sometimes do get lost going down the rabbit hole. But I've never found a situation where not going down that rabbit hole does not help me in my day to day job, especially in AI where everything is moving so fast.

• 41:15Claire Vo?

  Yeah. And this is just to make everybody feel comfortable. This is not just a beginner thing. And I find myself doing this with GPT-5.4, which is like a powerhouse model and also like talking to the most esoteric senior software engineer you've ever met, where it like explains its plans in these very technical terms. And I'm like, dude, just like, explain to me what you're doing in number one. Tell me in plain language. I do not need the technical details, like just tell me in plain language. And again, it comes from this curiosity of I want to make sure I understand the fundamental concepts of what you're talking about. And I want to make sure I'm learning both my code base and general principles as we go. And so I do think that curiosity mindset, no matter

• 41:58Claire Vo?

  what your seniority level is, your experience with technology, you can always learn, learn something better. Okay, my last question before we get you out of here. When AI is not giving you the right answer, it's giving you AI slop that it wants to email to your customer. What is your prompting technique?

• 42:18Speaker 00

  I'll say one thing first off the bat is like, I'm very relentless when it comes to getting the right answer from Claude or from AI. Like I treat it like my entry level analyst, you know, to that I can just like throw a billion questions. I, because I used to be an analyst and I come from the world where like you were just expected to crank. And so I'm relentless when it comes to asking AI to do things for me, especially when it comes to answering customer questions. I think the one prompt strategy I use is like, I mean, you've probably heard versions of this before, which is like, you know, my customer will, you know, churn if I don't get this right. Or, you know, like, my, my quota is dependent on getting this,

• 43:01Speaker 00

  this thing done. So those are kind of ways I've approached it, but those are like half answers. I think the, the real answer, and this goes back to curiosity thing is like, and think hard, think harder about why you're giving me this answer. Think hard about why this is right. And so in Claude code, there's actually like this think hard, think harder paradigm of like, how much reasoning it does to come to the answer. And so it's just going one step deeper and saying like, you gave me the answer, tell me why you think this is the, why this is the right answer, and give me the sources for what provided you with this reasoning. So I think going that one extra step, especially for those questions where you're

• 43:43Speaker 00

  like, not quite sure if it's the right answer. And like, you're reading the code and it kind of makes sense, but just going that one extra query to make sure you're getting the right response will sometimes, you know, give you new insights about your code base that on your product that you haven't thought about before.

• 43:59Claire Vo?

  Okay, I like the practical, like force the enhanced reasoning, think hard, think harder. I don't want people to miss. You tell people you're going to miss quote. I mean, like, let's give Claude code a little spiff every time it answers a question correctly. You've got to split your quota with Claude code. That's really what we need to do. And say, look, I'll give you a point on this deal if we can answer this question. Very, very funny. And I think today it'll be live by the time this episode goes live. Stripe just released this like payments protocol so you can pay your agents. So you can, you know, toss it a couple agent bucks or whatever.

• 44:39Speaker 00

  Yeah, because you better answers the more bucks you give it or something.

• 44:42Claire Vo?

  That's exactly it. Claude, coin operated Claude. That's going to be my new, my new skill. Well, Al, this was great. Where can we find you and how can we be helpful?

• 44:51Speaker 00

  I'm on LinkedIn, Al Chen on LinkedIn at Galileo. And check out Galileo if you're building agentic applications. Also, I think number one thing for me is my team is actively hiring field engineers. So if you want to work in post-sales, pre-sales, forward deployment engineering, we have a bunch of open roles. I would love to have you join the team if this is of interest.

• 45:14Claire Vo?

  Amazing. Thanks for joining How I AI.

• 45:17Speaker 00

  Thank you so much.

• 45:19Claire Vo?

  Thanks so much for watching. If you enjoyed this show, please like and subscribe here on YouTube, or even better, leave us a comment with your thoughts. You can also find this podcast on Apple Podcasts, Spotify, or your favorite podcast app. Please consider leaving us a rating and review, which will help others find the show. You can see all our episodes and learn more about the show at howiaipod.com. See you next time.