Evals, error analysis, and better prompts: A systematic approach to improving your AI products | Hamel Husain (ML engineer)

Hamel Husain, an AI consultant and educator, shares his systematic approach to improving AI product quality through error analysis, evaluation frameworks, and prompt engineering. In this episode, he demonstrates how product teams can move beyond “vibe checking” their AI systems to implement data-driven quality improvement processes that identify and fix the most common errors. Using real examples from client work with Nurture Boss (an AI assistant for property managers), Hamel walks through practical techniques that product managers can implement immediately to dramatically improve their AI products. What you’ll learn: 1. A step-by-step error analysis framework that helps identify and categorize the most common AI failures in your product 2. How to create custom annotation systems that make reviewing AI conversations faster and more insightful 3. Why binary evaluations (pass/fail) are more useful than arbitrary quality scores for measuring AI performance 4. Techniques for validating your LLM judges to ensure they align with human quality expectations 5. A practical approach to prioritizing fixes based on frequency counting rather than intuition 6. Why looking at real user conversations (not just ideal test cases) is critical for understanding AI product failures 7. How to build a comprehensive quality system that spans from manual review to automated evaluation — Brought to you by: GoFundMe Giving Funds http://gofundme.com/howiai—One account. Zero hassle: https://gofundme.com/howiai Persona https://withpersona.com/lp/howiai—Trusted identity verification for any use case: https://withpersona.com/lp/howiai — Where to find Hamel Husain: Website: https://hamel.dev/ Twitter: https://twitter.com/HamelHusain Course: https://maven.com/parlance-labs/evals GitHub: https://github.com/hamelsmu — 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 — In this episode, we cover: (00:00) Introduction to Hamel Husain (03:05) The fundamentals: why data analysis is critical for AI products (06:58) Understanding traces and examining real user interactions (13:35) Error analysis: a systematic approach to finding AI failures (17:40) Creating custom annotation systems for faster review (22:23) The impact of this process (25:15) Different types of evaluations (29:30) LLM-as-a-Judge (33:58) Improving prompts and system instructions (38:15) Analyzing agent workflows (40:38) Hamel’s personal AI tools and workflows (48:02) Lighting round and final thoughts — Tools referenced: • Claude: https://claude.ai/ • Braintrust: https://www.braintrust.dev/docs/start • Phoenix: https://phoenix.arize.com/ • AI Studio: https://aistudio.google.com/ • ChatGPT: https://chat.openai.com/ • Gemini: https://gemini.google.com/ — Other references: • Who Validates the Validators? Aligning LLM-Assisted Evaluation of LLM Outputs with Human Preferences: https://dl.acm.org/doi/10.1145/3654777.3676450 • Nurture Boss: https://nurtureboss.io https://nurtureboss.io/ • Rechat: https://rechat.com/ • Your AI Product Needs Evals: https://hamel.dev/blog/posts/evals/ • A Field Guide to Rapidly Improving AI Products: https://hamel.dev/blog/posts/field-guide/ • Creating a LLM-as-a-Judge That Drives Business Results: https://hamel.dev/blog/posts/llm-judge/ • Lenny’s List on Maven: https://maven.com/lenny — 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.

• 0:00Claire Vo?

  What are the fundamental concepts folks need to know of getting to higher quality products?

• 0:04Speaker 01

  The most important thing is looking at data. Looking at data has always been a thing, even before AI. There's just a little bit of a twist on it for AI, but really the same thing applies.

• 0:14Claire Vo?

  When you see a real user input like this, you actually look at what users are prompting your AI with. You realize it's very vague.

• 0:21Speaker 01

  Absolutely. That's the whole interesting bit. It's like, once you see that people are talking like that, you might actually want to simulate stuff that looks like that because if that's the real distribution of the data or that's what the real world looks like.

• 0:33Claire Vo?

  I'm sure our listeners expect some like magical system that does this automatically. And you're like, no, man, just spend three hours of your afternoon, go through, read some of these chats, look at some of them with your human eyes, put one sentence notes on all of them and then run a quick categorization exercise and get to work. And you see this have actual real impact on quality and reducing these errors?

• 0:55Speaker 01

  Yeah, it has immense quality. It's so powerful that some of my clients are so happy with just this process that they're like, that's great, Hamel, we're done. And I'm like, no, wait, we can do more.

• 1:09Claire Vo?

  Welcome back to How I AI. I'm Clairveaux, product leader and AI obsessive here on a mission to help you build better with these new tools. Today I have such an educational episode for people like me that are building AI products. We have Hamel Hussain, who is going to demystify debugging errors in your AI product, writing good evals and show us how he runs his entire business using Claude and a GitHub repo. Let's get to it. This episode is brought to you by GoFundMe Giving Funds, the zero fee DAF. I want to tell you about a new product GoFundMe has launched called Giving Funds, a smarter, easier way to give, especially during tax season, which is basically here.

• 1:52Claire Vo?

  GoFundMe Giving Funds is the DAF or donor advice fund from the world's number one giving platform, trusted by 200 million people. It's basically your own mini foundation without the lawyers or admin costs. You contribute money or appreciated assets, get the Tax deduction right away, potentially reduce capital gains and then decide later where to donate. From 1.4 million nonprofits. There are zero admin or asset fees and while the money sits there, you can invest and grow it tax-free. So you have more to give later. All from one simple hub with one clean tax receipt. Lock in your deduction now and decide where to give later. Perfect for tax season.

• 2:36Claire Vo?

  Join the GoFundMe community of 200 million and start saving money on your tax bill. All while helping the causes you care about the most. Start your giving fund today in just minutes at gofundme.com/HowIAI. We'll even cover the DAF pay fees if you transfer your existing DAF over. That's gofundme.com/HowIAI to start your giving fund. Hamel, I'm really excited for this particular episode because I have been building products for a very long time. And this has been one of a few times in my career where

• 3:19Claire Vo?

  the how and what of products that I'm building are so different than what I've built in the past. They're technically different, they're different from a user experience perspective. And then they have, they have these non-deterministic models on the backend that I'm somehow, as a product leader, responsible for making output, high-quality, consistent, reliable, interesting user experiences. And it's such a challenging problem and what I love about what you're going to show us today is how to approach that systematically, that quality of product building in an AI world systematically.

• 3:58Claire Vo?

  And how you use different techniques to get AI products, which are new to all of us, from good to great.

• 4:06Speaker 01

  Yeah, happy to be here. Excited to talk about it.

• 4:09Claire Vo?

  So, you know, this is such a new thing for product managers. I'm curious if you could start with the fundamentals. What are the fundamental concepts or things that you think folks building AI products really need to know about the process of getting to higher quality products? And then I know you're going to show us a couple of examples of how to do that.

• 4:29Speaker 01

  So the fundamentals really come down to the most important thing is looking at data. And I believe from working with many product managers in the past is looking at data has always been a thing, like even before AI, you know, like I'm pretty sure that product managers that can like write a little bit of SQL are okay with spreadsheets, looking at numbers, looking at metrics. You know, that feels like it's kind of table stakes for being a good product manager nowadays. And so there's just a little bit of a twist on it for AI, but really the same thing applies. And it's just like, okay, how do you do that for AI? And that's, that's what we teach and that's what I'm going to show you today.

• 5:11Claire Vo?

  Great, and I cannot agree more. I think one of the most transformational skills I learned as a young baby chicken product manager was being able to write SQL and actually do my own data analysis and exploration. But I think the surface area is so broad now with AI and the data is different. So why don't you show us what we should be looking at when we're building these AI products?

• 5:33Speaker 01

  Yeah, so let me share my screen a bit. Let me give you some background first. So this is one of my clients. The name of the company is called Nurture Boss. And as you can see, it's an AI assistant for apartment managers or property managers. And really like, you know, you can kind of get an idea from their website, which I'm showing right now. You know, it's a virtual leasing assistant. So, you know, they help with the whole top of funnel of like helping set up appointments, helping prospective residents like find their apartments, setting up appointments, questions about rent, so on and so forth, kind of like trying to reduce the toil of property managers,

• 6:18Speaker 01

  still having humans in the loop. And so when they came to me, they had already prototyped something out, you know, kind of vibe checking it, just like everyone does, and put everything together, but they wanted to know, like, okay, how do we actually make it work well? Because the AI fails in weird ways and doesn't always do the right thing. But it feels like, okay, every time you fix a prompt, we're not really sure. Like, maybe we're breaking something else. Or is it really improving things as a whole? We don't really know, we're just guessing, we're just kind of like looking at it. And just getting vibes. And that is a very uncomfortable feeling of trying to scale a product. Okay, so the first thing that I'll jump right into is this idea of traces.

• 7:04Speaker 01

  So traces are this concept from engineering, but doesn't have to be scary. They're basically like, and it's very topical for AI because with AI usually have many different events. Or especially like for a chatbot, you have multi-turn conversations where you're going back and forth with an AI. There might be retrieval of information. They might be calling some tools, external tools, internal tools, so on and so forth. And so you want to log these traces. And there's, there's many different ways to go about it. But just to kind of show you exactly what happened at Nurtureboss, let's go into what that looks like. So this is a platform called BrainTrust.

• 7:49Speaker 01

  There's a lot of them. This is one called Phoenix, which is like the same exact data in here. It doesn't really matter, you can see like they're both the same, right? Like, so what we have here, let me just go into a single trace. So this is what I would call a trace. I can make this bigger so you can see in a full screen. And you can see what an AI interaction looks like in this product. So you have, okay, the system prompt, you are an AI assistant working as a leasing team member at some apartment. These are all fictitious because these have all been scrubbed for PII stuff. You know, your primary role is to respond to text messages. So this is receiving text messages.

• 8:31Speaker 01

  Okay, and you have a whole host of rules like respond, you know, provide accurate information, answer any question for residents, do the following, you know, provide this website. For example, if you had asked for a rental application, provide this on and so forth, all these rules, right? And this is a real user saying, hello, there's what's up to four month rent. I don't even know what that means.

• 8:56Claire Vo?

  I got you. I got you. Let me read it. Hello, hello there. What's up? Two, four month. I thought I had it. I thought I had it.

• 9:07Speaker 01

  Yeah. It's unclear, but okay. I mean, like, it's fine. This is real. This is the real world. These are real traces. So, um, you know, and then there's a, there's a tool call here. Get communities information. It's calling this tool, this internal tool. And the tool call result comes back with this information. And this is all hidden from the user. The user is not seeing this tool call result.

• 9:34Speaker 01

  You're like, okay, here's the information you can use about the community, blah, blah. It's not even sure like this is the right tool call. We'll get to that in a moment. And then the assistant goes, hello, we are currently offering up to. So this is like back to the user. This is what the AI responds to the user with. Hello, we are currently offering up to eight weeks rent free as a special promotion. Please note the applicable lease specials and concessions can vary, blah, blah. Okay. So like, is this, and I have a cheat sheet for myself about what is actually right and wrong. Okay, so like, the comment here is the user is probably asking about lease terms and stuff like that, not about specials.

• 10:19Speaker 01

  So like, it's not really clear like this is the right. This is not like what we want. And this is so realistic, right? Like, everyone has experienced AI like this. It's like, it's kind of, it's being helpful, but it's not really doing what you want to. And it's actually pretty challenging because it's not really clear what the user wants. You could go in a lot of different directions.

• 10:37Claire Vo?

  You know, when I'm testing my own AI, this is such an eye-opening example. Because when I'm testing my own AI, I ask it good questions and I spell correctly, and I'm very clear. But when you see a real user input like this, you actually look at what users are prompting your AI with. You realize it's very vague. They say stuff like, "What's up?" The question, there's no clear question. And so I really do think, looking at real user data, kind of can get a developer or PM out of their own mind on how they think users are going to interact with the system.

• 11:13Speaker 01

  Absolutely. It's very critical that you do this. And so, now, you might not have this data. And I just jumped right into a real example just to set things off. And we can go into all these different rabbit holes. Like, what if you don't have data and stuff? But I just want to like, round it and like, okay. So to set the stage, like, this is kind of one foundation. It's like, you have to have data. There's different ways to get it. One is you can log it from your real system, and you have these things to look at. Another way is like, okay, you can have synthetic data, where you sort of generate with an LLM. You can generate questions like this. You know, hello, what's, you know. It might be hard to generate stuff that looks like that because I don't even know, we don't know what it means.

• 11:57Speaker 01

  And probably an LLM won't generate stuff like that. But that's the whole interesting bit. It's like, once you see that people are talking like that, you might actually want to simulate stuff that looks like that. Because if that's what, if that's the real distribution of the data, or that's what the real world looks like, you might want to challenge your LLM or your AI system appropriately. Okay, so let's step back here. So you have the system. It's doing, it's like, there's stuff like this happening. We can look at another trace if you want, just to kind of get an idea. And this is, you know, this is not pre-scripted. I didn't memorize what's going on in these traces. We're just looking at them naturally.

• 12:38Speaker 01

  So this is something, this is another apartment complex, Meadowbrook Apartments, same idea, so we won't read the whole system prompt again. Okay, so we'll scroll down here, let's get to what the user is asking. Walk-in T O R. So this must be another text message situation. And the assistant says, our team tries their best to accommodate walk-ins. Me get you. No, that's hilarious. Like, I don't, why is LM that surprising? Like, why is it saying me get you to someone who can help? Maybe he's trying to mimic the, uh, the, uh,

• 13:12Claire Vo?

  The tone.

• 13:12Speaker 01

  The user somehow. Um, and then it does like, uh, yes. And then, okay, great. So it seems like this one maybe is okay. Um, let's see what we end up annotating. Uh, yeah, we said this one is okay. There's, there's some metadata down here about our labels, which we'll talk about next. But yeah, you can, so you can see like, this is a real system. There's many different things that can happen here. So the question becomes like, okay, so we talked about this, like, writing SQL and data. But like, how do you take that same mindset to this? Like, what do you even do with this? Right? You have this like crazy, like, interactions, like, how do you analyze this? Without, without getting stuck?

• 13:55Speaker 01

  Cause like, this seems like, um, intractable, right? At first pass.

• 14:00Claire Vo?

  No, I, I was just thinking, I was like, what is the SQL query I write to get like the first prompt? And like, how do you query for, give me all the first prompts that include typos? Like, give me the, all the first prompts that are ambiguous questions. It just feels almost insurmountable. And then, you know, you showed us two examples and it's two of probably thousands and thousands and thousands. So going through it manually is probably not super scalable. So I'm curious, what is the systematic kind of solution here?

• 14:30Speaker 01

  Okay. So the systematic solution is called, something called error, error analysis. So error analysis just means, It's kind of a counterintuitive process that's extremely effective and it's dumb, but it's accessible to everybody and it works. And it's not something that I made up. It's been around in machine learning for a really long time because actually, machine learning has the same problem like before like generative AI. We had these stochastic systems that can do like a whole number of things. And like, how do you actually like analyze that and like figure out like what's going wrong and improve it? So error analysis has two steps.

• 15:12Speaker 01

  The first step is writing notes and it's called open coding. And it's basically like journaling what is wrong. So if we go back to like that, that other trace that we saw, so let me just go back to it like the first one. We would step into this trace and we would say, okay, like every, every observability tool has their own, let's say, different ways to take notes. You know, I already have a note in here. Assistant should have asked follow up questions about, you know, about the question, what's up with four month rent? Because it's unclear user intent. I just was writing notes about what is going on.

• 15:51Claire Vo?

  Yep.

• 15:52Speaker 01

  And you do that for like a hundred traces. Randomly sample a hundred traces and you do that. And you stop at the most upstream error you find. So you read this and you see what's going on and you're like, hmm. Okay, the user intent seems like we didn't do a good job of like clarifying that what the hell that they're, they need.

• 16:11Claire Vo?

  Yeah.

• 16:12Speaker 01

  And so I think that's the most upstream problem in this sequence of events. So I'm going to go ahead and just write that as a note.

• 16:19Claire Vo?

  Yeah. And, and you say focus on the most upstream problem because you presume that if you can get early intent, early kind of clarity, correctness, right? The rest of the system is more likely to be correct downstream.

• 16:33Speaker 01

  Yeah, because it's causal in nature. So as we have the sequence of events, whether it's like user prompts, tool calls, retrieval for RAG, whatever it may be, any error at any point along the chain. You know, like, will cause downstream problems. And so to simplify our lives for this purposes of error analysis, it's heuristic. You know, eventually you do want to care about the different errors and different downstream. But when you're starting out, just focus on the upstream error. Because we're trying to make it tractable. And this is like the way that you're gonna get results fast. So basically, what you do is you go through and you collect a bunch of notes.

• 17:14Speaker 01

  And then what you do is you can take these notes and you can like download them or whatever. And you can categorize those notes. And you can even put these notes into like chat GPT. It's like, hey, here's all my notes. Like, can you bucket these into categories? And you kind of have to go back and forth with it a little bit like, hey, these are my notes. These are the categories. Um, I think like you're missing a category, whatever. Now with nurture boss, what we ended up doing is we actually made one of the things that we highly recommend, a lot of people think about is to make your own custom annotation tool. Like there's, you see this is here in brain trust and it's also here in arise phoenix. They're very similar.

• 17:59Speaker 01

  You can see this is a very similar looking UI. And you have, they even called it error analysis here. And you can like add your notes, like, you know, whatever. And you can save those notes and same thing. If you're gonna be looking at a lot of data, you don't want to slow yourself down. And you want to be able to have like very human readable sort of, you know, output. And sometimes like this markdown stuff is like not that readable. And you want to make sure that, okay, like it makes sense to you and you can fly through it as fast as possible. So, um, you know, it's really easy to vibe code this stuff.

• 18:41Speaker 01

  Because ultimately what you're doing is like showing data. So when the, in the nurture boss situation.

• 18:48Speaker 01

  So as you might have gathered, like they have multiple channels that customers can contact them on. They have like text message, which we saw. They have email. They have a chat bot on the website, so on and so forth. So they just wanted something they could like navigate faster. It's just like vibe coded essentially. I mean, they have, we were developers, but you know, we're using AI in our process and do this very fast. Is okay, like what channel is the trace from? And then like some other filters about like, hey, did we already annotate this or not? And then just kind of have some statistics at the top. You know, this is like what the annotation like looks like. It's kind of very similar, but just like dialed into what we wanted. And like, you know, we just took notes.

• 19:32Speaker 01

  And then what for nurture boss, what we did is okay, we had an automated process that would summarize, like categorize those notes into like, what are the biggest issues? And then we would just something very simple like counting. Counting is always powerful. As you know, as a product manager, you can go into a system to see your, you experience like writing SQL queries. Like you know how powerful counting is. Counting remains powerful. And so you can count these issues, right? So, so like, okay, for nurture boss, I don't know if you can see my screen or if it's too small.

• 20:09Claire Vo?

  Yeah, that's great.

• 20:10Speaker 01

  Is, okay, what are the most, what are the biggest issues after doing that error analysis exercise? Which only took, you know, a few hours.

• 20:19Claire Vo?

  Yeah.

• 20:20Speaker 01

  It's like, okay, we're having a lot of transfer and handoff issues. We're trying to transfer the, the customer to a human. We're having a lot of tour scheduling issues. So like they're trying to schedule a tour, but like a reschedule tours. In this case, we found that like someone's asking to reschedule. There is no rescheduled tour. But like the AI doesn't know that. It just keeps scheduling more tours, which is bad.

• 20:45Speaker 01

  You know, follow up. So, you know, AI not following up when the user has a question. You know, sometimes incorrect information provided. Okay. So like you see like these are kind of the count. And now we have, now we're not lost. Now we know what we should be working on. We know, okay, you know what? We should fix this like transfer and handoff issue and this tour scheduling issue. We have confidence, like, you know what? Like we were not paralyzed anymore. We know. Okay, this is what we need to fix it on our AI.

• 21:18Claire Vo?

  This episode is brought to you by Persona, the B2B identity platform helping product, fraud, and trust and safety teams protect what they're building in an AI-first world. In 2024, bot traffic officially surpassed human activity online. And with AI agents projected to drive nearly 90% of all traffic by the end of the decade, it's clear that most of the internet won't be human for much longer. That's why trust and safety matters more than ever. Whether you're building a next-gen AI product or launching a new digital platform, Persona helps ensure it's real humans, not bots or bad actors, accessing your tools. With Persona's building blocks, you can verify users, fight fraud, and meet compliance requirements, all through identity flows tailored

• 22:03Claire Vo?

  to your product and risk needs. You may have already seen Persona in action if you've verified your LinkedIn profile or signed up for an Etsy account. It powers identity for the internet's most trusted platforms. And now, it can power yours, too. Visit withpersona.com/HowIAI to learn more. I love this. Just to recap, so you're taking these traces of these real conversations. And you don't even have to read all of it. You have to read till you hit a snag, right? To hit an obvious, sort of like incorrect or high-friction part of the experience. You have vibe-coded an app that makes it really easy for the team, generally, to go in, annotate these, rate them, sort of like,

• 22:48Claire Vo?

  good quality, bad quality, automatically categorize them, count them. And then you have a prioritized list, and you're like, here are the problems that I need to go solve. And what I love about this is, you know, I'm sure our listeners expect some, like, magical system that does this automatically. And you're like, no, man. Just spend three hours of your afternoon, go through, read some of these chats, look at some of them with your human eyes, put one-sentence notes on all of them, and then run a quick categorization exercise and get to work. And you see this have actual, real impact on quality and reducing these errors?

• 23:26Speaker 01

  Yeah, it has an immense quality. It's so powerful that some of my clients are so happy with just this process. That they're like, that's great, Hamel, we're done. And I'm like, no, wait, like, we can do more. You know, you've paid for more, like, you know, whatever. They're like, no, this is so great. Like, I just feel like I know what to do. And so they find so much value in this like process that. And it is like very important. This is something that no one talks about. The people, when you talk about evals, like, well, how do you write an eval? What eval do you do? What tool should you use? Before you get into all that stuff, you need to have some grounding in like what eval you should even write. Because there's infinite evals.

• 24:11Speaker 01

  So like in this case, we were right, we wrote an eval about tour scheduling issues. And we wrote an eval about transfer handoff issues. And we felt really good about that because we knew that like that is a real problem. And we, we knew how to write the eval because like we saw the error. And we knew how to find data to test that eval. Because again, we already tagged it and we saw that error, which is exactly the way you want to do it.

• 24:33Claire Vo?

  Yeah, and what I also like about this is it does take the burden off your users. I mean, so many people try to collect this data by like putting a little thumbs up and thumbs down or little comments, like I even have that on parts of my product. And yes, it is useful, but it only gives you a sliver of the kind of self identified errors in the app. And users are highly tolerant of systems. And so sometimes those errors just don't get escalated by user. They'll either abandon or they'll just work through too many steps to get to the outcome that they want. They'll have a quality experience. And so I think just taking the burden on yourself and saying you're responsible for looking at the data. You can create simple ways to categorize it. And then you have a prioritized list.

• 25:16Claire Vo?

  Now, if your client is willing to go the next step and do something about this and write evals and fix prompts, what are your kind of next steps here? What's another example of where we're from here?

• 25:31Speaker 01

  I just want to talk about this for a minute. Like, okay, so this particular technique is so powerful. And I mean, people know about it. You know, so I actually recently did a training with OpenAI showing the people at OpenAI like, you know, how this works for domain-specific evals. If you want to learn more about like this, we had Jacob, the founder of NurtureBoss, like walk through like this whole process in like two minutes. So you can find it on this, on this page if you would like.

• 26:01Speaker 01

  Okay, so get to your question, like, what do you do now? Okay, so you have, like, you know, you've done your error analysis. You have like prioritized these things. So like now what do you do? So now you get into writing the evals. So now you have to decide like what kind of evals do you want? There's different kinds of evals. So there's reference-based evals, which is like you know what the right answer is. And maybe you can write some code. You don't need like an LLM to do the eval for you. Or if it's more subjective in nature, then you know, maybe like this transfer handoff issue,

• 26:42Speaker 01

  maybe it's more subjective in nature, then you need an LLM judge. And so what you can do is you can start to write those evals. And so I had this blog post here about evals in general. So there's this diagram. It's really hard to put this whole thing into a diagram, honestly. But because, you know, it can be, it's kind of, it's not, it's nonlinear process. But really what you want to do is, okay, we already covered like logging traces. And there's two different kinds of, but there's different kinds of evaluators or evaluations. There's the kind of like unit tests, which is like, well, I would say like code-based evals.

• 27:26Speaker 01

  And then there's like models, so like LLMs. You know, code-based evals, so like, you know, for example, what is, what are the kinds of things that be good for code-based evals? Like, okay, if you have like user IDs showing up in the response or something like that, okay, you can test for that in code.

• 27:45Claire Vo?

  I have to say you're saving my life here because I was thinking, what is one of these unit tests I need to write? And that is exactly one of them, which is my tool calls need to be IDs and users definitely do not. So that's a, that's a great example of one for anybody that's writing a chatbot that does a lot of kind of tool calling.

• 28:02Speaker 01

  Yeah, because they can show up by accident. Like me, you might have the UID in the system prompt and it inadvertently shows up in output for some reason or another and you don't want that. Okay, you want to write these tests with, no matter what kinds of tests you write, you want to create test cases. And sometimes you can gather those from your traces. Sometimes, sometimes you might want to generate synthetic data. And so, you know, this is like a prompt for a different real estate agent assistant called Reachout, which is for residential real estate. And this is kind of like a simplified version of your prompt, write 50 different instructions that a real estate agent can give to their

• 28:45Speaker 01

  assistant. It creates contacts on their CRM. Contact details can include name, phone, email, whatever. And basically, you know, it can generate synthetic inputs to a system that then you can then log traces from. I'm going to jump around a little bit, so we'll kind of come back to that. Okay, we already covered logging traces. You know, this is another like custom log annotation thing yet again, because we, you know, really emphasize this, that it's really important to remove all friction doing this. So I won't linger on this too much. And basically, you know, one kind of thing you want to do is like,

• 29:26Speaker 01

  okay, if you're using LLM as a judge or anything else, what you want to do is, so one thing that's usually skipped when we talk about LLM as a judge, is like people are just using LLM as a judge off the shelf. Like they're like writing a prompt, they're saying, okay, judge it, and then reporting that. Let me actually go to a different blog post that is a little bit better for LLM judge.

• 29:53Speaker 01

  Which is this one. Okay, so LM as a judge. So you often see sometimes in LM eval land, like a dashboard that looks like this. Helpfulness, truthfulness, conciseness score, tone, whatever. What the hell does that mean? Does anyone know what that means? Nobody knows. No one understands concretely. Like if the helpfulness score is 4.2 and it goes to 4.7, like, do you really know like what's wrong? What changes? No. And so there's a lot of guidance in how to create LM as a judge. It's probably too much for this podcast to like tell you all of the things. And this blog post is quite long, like enumerating like how to do it correctly.

• 30:39Speaker 01

  But the main things you need to keep in mind is like, one, you need to have binary outputs. Like, is it good or bad? For a specific problem. So for like, you know, the handoff problem for nurture boss, like, okay, was there a problem or not? And you want specific evaluators for specific problems. Number two is like, you want to, you need to hand label some data, which you already kind of do in error analysis. And you want to compare the judge to the hand label data so that you can trust the judge. The last thing you want to do is like throw up a judge on the dashboard like this. And then like,

• 31:16Speaker 01

  people don't know if they can trust it. And the worst thing you do as a product manager is like, start showing people evals. And then at some point, the people's perception of the product or their experience of the project doesn't, doesn't match the eval. So like, hey, like it's, it's broken, but the evals are showing that it's good. And that's a lot. That's the moment like people lose trust in you. And then they'll like, it's gonna be really hard to regain that trust. And so the way that you make sure you can trust these automated LM evals is to, you know, measure sort of agreement with these hand labels.

• 31:56Claire Vo?

  Yep.

• 31:58Claire Vo?

  So what I'm hearing from you in terms of LLM as a judge is these general buckets with arbitrary ratings against them, not useful and will often work against you. You want to write specific binary outcome evals for specific tasks. So you want a set of evals that are like, does this get scheduled correctly? Yes or no? And so you're making a list of evals that the LLM as a judge is evaluating that gives you a pass, fail, or yes, no, true, false, binary outcome. Very simple. And then you're doing the additional layer of work of validating that the eval itself is valid by actually looking at that outcome and saying,

• 32:40Claire Vo?

  do I actually agree with this LLM as a judge evaluation of the quality of this output? And that those steps together are going to give you a much more comprehensive view of how your product's performing. And then, and then that, that second layer of human evaluation, it's going to give you more confidence that either your LLM as judge is good and is evaluating your outputs correctly, or you actually need to tune that judge itself to get to higher quality evaluations. Is that kind of the summary of what you

• 33:13Speaker 01

  Yes. And the thing that's really important is like, it's really difficult to write any LLM judge prompt if you don't do this because the research shows, and there's some research that my co-instructor for the course that I'm teaching, there's a paper called, Who Validates the Validators? And the research shows that people are really bad at writing specifications or requirements until they need to react to what an LLM is doing to clarify and help them externalize what they're, what they want. And it's like only going through this process of sort of, okay, writing detailed notes and critiquing things that you can then like start refining the LLM judge.

• 33:59Claire Vo?

  Great. And so we've, we've covered sort of traces and errors, annotation. You have kind of how to build unit tests that are automated tests. Of course, you're looking at it manually. You're doing LM as judge the correct way. Now tell me, I've identified all these problems. I have these evals that give me data. How do I write a good prompt? Like, are there, are there some techniques? Or, you know, what do I, what do I do? Are there things that you've found consistently in the next step of improving your system instructions? Improving your tools where you actually have to go solve these problems are, are effective?

• 34:39Speaker 01

  Yeah. So when you get to like the errors that you have, so like, you know, you're going to use these evals and you're going to deploy it at scale. Okay, it's like, you're not looking at all your data. You, you're looking at a sample of data.

• 34:54Speaker 01

  And you're going to score your LM as a judge against like a sample of label data. And you're going to deploy that at scale. And you're going to like, look at where are there errors? And it's pretty like, you know, you have to make a judgment call on like, how do, how do you improve your system based on the errors you're finding? Like, is it a retrieval problem? Is it a prompting issue? Is it, should you be putting more examples into prompt? And, you know, this is not really a silver bullet there, I would say.

• 35:29Speaker 01

  You know, retrieval is its own sort of beast. It tends to like, retrieval tends to be the Achilles heel of a lot of AI products. You know, where things tend to go wrong. But sometimes, yeah, it's just like, especially in the beginning, you're going to find a lot of low hanging fruits. Like, for example, in nurture boss, the system prop didn't contain today's date. So when the person said, hey, can you do a schedule for tomorrow? AI had no idea what like, we don't know what tomorrow is, but didn't, didn't tell the user that, right? We just guessed. So like, you know, that's really obvious. So there'll be like obvious things you can fix. And then there's like lesser obvious things you can fix. You could try like prompt engineering.

• 36:13Speaker 01

  So there's a spectrum of like, okay, prompt engineering all the way to like fine tuning. Most people shouldn't get into fine tuning. I will say that if you do all this eval stuff, fine tuning is basically free. Because you have all this infrastructure set up to do all these measurements and curate data, like high signal data that is difficult. And that difficult data, that those difficult examples where your AI is not getting right, that's exactly the stuff you want to fine tune on. That's like the very high value stuff for fine tuning. So, um, and yeah, fine tuning is not so hard. In the rechat case, we had to do fine tuning to get the extra mile.

• 36:54Speaker 01

  Um, but in most cases, like it's prompt engineering, there's no magic prompt engineering tricks. It's really like, I would say, there's a lot of experimentation that you should engage in.

• 37:04Claire Vo?

  Well, and one of the things that I found so interesting as an AI builder that comes from a software engineering background is now I have a natural language surface for bugs in terms of my system instructions and prompts. And I had this experience recently on chat PRD where we were really having a hard time with tool calling. Like one of our tools just was intermittently not being called no matter what the user would say, and it was really hard to pin down. And we have this, you know, monster system prompt and I went through and there was like two words in the prompt that were just incorrect. They were incorrect. It was about U U IDs, but it was like incorrect. And as soon as I deleted those two words, which had just been, you know, typed in by somebody and pushed in the repo, blah, blah, uh, our quality of that tool calling shot

• 37:49Claire Vo?

  right up. And so I just have to, you know, we have to, as product people, as engineers, start thinking of the full surface area of our product. And it's not the construction of the agent or the chat bot itself. It really goes down into what words are going in and out of your system. And it's a complicated surface area to debug and keep track of because it's unstructured, but it's super high impact in my experience.

• 38:14Speaker 01

  Yeah, definitely. You know, when it comes to tool calls, actually, let me show you one thing that always comes up is people wonder like, how do you evaluate agents? Because like, you know, there's so many different handoffs. Like, how do you actually like do it in real life? So let me see if I can share that. Okay, so I'm sharing like, um,

• 38:39Speaker 01

  The book that we give students in our class. Let me go to the table of content. So there's all these different areas. We'll kind of skim towards the agent part of it.

• 38:51Speaker 01

  So there's like analytical tools you can use for everything. You know, for agents, you can build these transition matrices. So going from one step to the other, where are the errors located? In like what agent handoffs or what steps being handed off to what other steps? So like in this case, okay, we have this like generate SQL to execution SQL. That's where a lot of this like errors are happening. And then you can like, then you can narrow it down. So as you get more advanced into evals, it's a very deep subject. You, there's a lot of analytical tools you can use to kind of go about things.

• 39:30Speaker 01

  It is very interesting, like as a product manager, you can get really far with AI assisted notebooks.

• 39:38Claire Vo?

  Yeah, what I was going to say about this from a product manager perspective is this is really put from the frame of errors and evals. But even just analytics for agentic systems, figuring out what your users are trying to do. I haven't thought of this idea of actually mapping out the different conversation to tool or tool to tool handoffs. And even if all of this was working effectively, a product manager's ability to see the data of its agent's behavior from a tool to tool handoff perspective and really identify like where are users trying to get value out of the system also can do things like drive roadmap ideas.

• 40:20Claire Vo?

  Right? If you're seeing, okay, people are just writing SQL, executing SQL. Like we need to dig into what other things around that could we build for users that are interesting? So I like it from the error perspective. I also like it just from the product discovery perspective.

• 40:34Speaker 01

  Yeah, definitely. That's, that's very true. Yeah, I like that.

• 40:38Claire Vo?

  Okay, so you've shown us how to, the other thing that I like that you've shown us. Is that there's no way to do this than just do it? Like I people want these tricks. They want some hack. They want some off the shelf solution. And you're saying like, honestly, look at the data. Build yourself a solution if you have to. Validate it yourself. Do the hard work. And if you do the hard work, you can actually create these leaps in product quality and experience. But right now, you just, you just got to look at the data and, and make some decisions and make things better. So I think this has been super illuminating in terms of helping people like me that are building AI products. Make them higher quality.

• 41:19Claire Vo?

  Let's spend just a couple of minutes on a totally different topic, which you are running this business. You're running a course. You are clearly an expert in AI. What tools are in your stack for kind of running your day to day life, or at least your business life?

• 41:34Speaker 01

  Yeah, so I do a lot of writing and I do a lot of communication with clients. And, you know, I also want to reduce my own toil. And so, um, let me share my screen again.

• 41:46Claire Vo?

  Yeah.

• 41:46Speaker 01

  It's probably easiest to show you Claude project. So I have all these Claude projects.

• 41:53Speaker 01

  So, okay, I have like one for copywriting. I have a legal assistant. I have consulting proposals. Consulting proposals is pretty interesting. So it's basically like an example of consulting proposals. It's, um, you know, I'm, so it's kind of funny. I have skill level, partner, palantir, is expert in generative AI, blah, blah. And, you know, I give it some instructions on the, on the other, like, let's say, proposals I have. And, you know, I have like this prompt, you know, whatever. Get to the point, write in short sentences, whatever. And basically I have a lot of examples and basically anytime I have a intake call with a client who wants a proposal.

• 42:36Speaker 01

  I give this the transcript and then it's made, it's basically almost ready. It's like, just need, it takes me about a minute to, to kind of edit it and get it going. So that's, that's proposals. You know, I have one for the course, which is like, you know, a lot of context about my course, which is like the entire book. Uh, I have an FAQ that's very like extensive that I've published. Um, there's all the transcripts, all the Discord messages, office hours, you know? And again, my prompt is like, hey, your job is to help course instructors to create standalone, interesting, epic queues. These are, this is like a writing prompt that I have everywhere.

• 43:17Claire Vo?

  Do not add lower words. Don't repeat yourself. Get to the point.

• 43:22Speaker 01

  Yeah, yeah, it's very, you have to really, you know?

• 43:26Speaker 01

  And so, okay, like, yeah, it's just, you know, this stuff here. You know, so there's like one for the course. There's, um, you know, there's one to help me create these things called lightning lessons, which is basically like, you know, this lead magnet.

• 43:45Speaker 01

  So there's all kinds of stuff like this.

• 43:48Claire Vo?

  I see you and I share a general counsel here.

• 43:52Speaker 01

  Oh, okay.

• 43:52Claire Vo?

  With Claude AI.

• 43:54Speaker 01

  Oh yeah, right, exactly. Yeah, there you go. So there's that and also have like, you know, my own software that I have.

• 44:02Claire Vo?

  Yeah.

• 44:02Speaker 01

  So I have, uh, let me see if I can find it.

• 44:07Speaker 01

  I mean, it's not, I'm not really advertising it, but I have like YouTube chapter creation. And then I basically have this thing that, um, will create blog posts like out of YouTube videos. So like, um, let me show you an example. So, uh, like this one, basically what I do is I take a YouTube video and it becomes an annotated presentation. So you don't have to watch the video.

• 44:31Claire Vo?

  Yep.

• 44:31Speaker 01

  Like you can just, especially if the video has slides, what it'll do is screenshot all the slides and then have a summary under each slide about what was said. So you can consume like a one hour presentation and like, you know, whatever, five minutes. Um, and that's really good because like, you know, I have, I teach a lot and I have a lot of content. And so I distribute notes. So all of that, so like a lot of that stuff, educational stuff is part of my workflow. And that this is used like this uses Gemini. Essentially what it does is it pulls the transcript. It pulls the video. I can put in the slides all at once and get have a lot of examples and I give it to it and it produces this.

• 45:12Claire Vo?

  Yeah, I've heard this in a couple of podcasts that we've done recently that folks really like Gemini for video information. In Jest seems to be the fan favorite for taking basically YouTube videos or other video content and turning it into text or other other applications that you can extract from that. So try the Gemini models for that, folks.

• 45:33Speaker 01

  Yeah, it's absolutely brilliant. It's amazing.

• 45:35Claire Vo?

  Cool. Okay, so you have Claude projects for every little part of your business. I love the proposal workflow. It's something that we, we, we folks that do enterprise sales could probably make, make some use out of. I'm about to start doing blog posts on all the how I AI podcasts. So maybe I will download your repo and give that a little spin. And then you're using Gemini models to extract out content and share it as, as templates. And then you have, oh, look at these prompts. We got a GitHub with prompts.

• 46:04Speaker 01

  Yeah, so I get GitHub with prompts. This one is private, but just to give you an idea, conceptually, like I, it's basically a mono repo of everything. The reason that is, is because I like to have Claude code, open hands, you name it. And basically what I say is because all these things are all interrelated, right? Like a lot of these projects, so like, you know, uh, this is my, my blog is in here. This is my blog, for example. This is that, that like YouTube thing I just showed you, this Hamel project. This is like something else that fetches Discord. This is about copywriting proposals, whatever. And I just point AI at this repo. And, you know, there's like Claude rules in here that says like, okay, what is this repo about?

• 46:49Speaker 01

  And like, where do you find stuff like, okay, you know, this is like, if you need to like, for writing, you should look here. You know, so on and so forth.

• 47:00Claire Vo?

  So my friend, you have buried the lead. Here because we could have done an entire episode on just this repo. What this makes me think of is, you know, five, five years ago, there is this big like note taking, second brain, where do you put all your information so you can have access to it forever? And I see this and my little engineering brain goes, obviously it should go in a repo. And it should be a combination of data sources, notes, articles, things that I've written, things that I like and prompts and tools to actually do something with that. So you have given me a personal project that I'm going to go work on in the next couple of days. Cause I think this is, this is how I, as somebody who lives with cursor or Claude code as sort of co-pilots

• 47:45Claire Vo?

  for everything I do, this is how I would want to organize my data and my prompts to be able to do something with it.

• 47:53Speaker 01

  Yeah, I don't want to be locked in, right? Like to any one provider. And so this is how I do that.

• 47:58Claire Vo?

  It's amazing. Okay. We might have to have you back to go through this thing in detail. This has been so great. Um, I have two lightning round questions for you and then I will get you out of here. I know you're a busy guy. My first question is, you know, a lot of what you showed us requires someone, a person to go through with their human eyes, read things and evaluate. And I'm curious, whose role do you think this is? Is this the product manager's role? Is it the engineer's role? Is it the subject matter expert's role? Who, who does this?

• 48:27Speaker 01

  I think the subject matter expert is very central. A lot of times the product manager is the subject matter export, you know, SME in a lot of organizations. Like they're kind of the person that everyone looks to for like the taste of like, hey, this is what should be happening with the user. So I would say a lot of times it is the product manager that should be doing that annotation.

• 48:51Speaker 01

  Now, when it gets into the analysis, it's really interesting. It would be good if a product manager, like the more you can do, the better. Just like the sequel and the stuff that you know about. At some point you do need, you'd probably need a data scientist when it gets advanced.

• 49:08Speaker 01

  But you know, the more you learn, the better and vice versa. The more data scientists learn more product skills, you know, it's going to be better. It's hard to predict like.

• 49:18Speaker 01

  You would, you know, there's always this tension or this kind of, okay, can we collapse roles? Can we collapse the product role in this like data scientist type AI role? I'm not sure. It's yet to be seen. I don't think so. There's a lot of surface area, actually. There's like, there's something called AI engineer. There's AI product manager. And there's not, there's also like still this data scientist aspect. So those three roles are still operating on this problem. And there's, there's definitely a lot of surface area for all of them, especially as you scale.

• 49:54Claire Vo?

  The, the one other thing that I would call out, or my hope is in addition to sort of like the technical building teams who are sort of proxies in, in my mind for the subject matter experts. So a lot of times the product manager is a proxy for like the leasing agent in this example. They understand that user, they understand what high quality is. But, you know, I would really love to see folks that are in operational or more functional roles come in and actually contribute to the quality of the products because you know what makes a good user experience. You know what makes a good leasing agent. You know how they should speak and what they should do. And I think there is an opportunity for folks to lean in and bring that expertise to bear in a way that scales across a company.

• 50:36Claire Vo?

  That if you're willing and brave to do it, I think product teams would welcome in kind of like non-technical colleagues into this process to add some more kind of user empathy and subject matter expertise.

• 50:49Speaker 01

  Yeah, definitely. Yeah. The more you can bring like the actual required taste in the product sense into the process, the more that, yeah, because that's essentially what you're doing when you're annotating.

• 51:00Claire Vo?

  Yep.

• 51:00Speaker 01

  And doing this error analysis and the error analysis is the foundation for everything.

• 51:04Claire Vo?

  Yep. Okay. And then my final question, ask everybody, I know you're very structured and you'll tell me you'll look at the data and then figure out exactly what to say. But you have to admit, sometimes AI is very frustrating and doesn't do what you want it to. Do you have any back pocket prompting techniques you use? Do you yell? Are you all caps? What, what's your, what's your strategy?

• 51:25Speaker 01

  AI has frustrated me the most is writing. Cause like writing, I don't want the writing to sound like AI.

• 51:32Claire Vo?

  Yeah.

• 51:32Speaker 01

  And it's hard, you know, that's the last thing you want in certain situations for your writing to sound like AI. And not that AI is like wrong, it's just that, yeah, you want to make sure your like flavor is coming across. And so, um, so one thing, one thing is like, okay, I showed you my writing prompt, a little bit of it. I can share it with you separately also is like, provide lots of examples, but then also take it step by step. So for writing, what I do is have it write an outline and then I have it write the first one or two sections and edit it very carefully. Now, one tip is use something like AI Studio that allows you to edit the output of

• 52:12Speaker 01

  what the L M is giving you. That's really important because like what that ends up doing is it creates examples for the L M in kind of right there.

• 52:22Claire Vo?

  Yeah, inline. Yep.

• 52:24Speaker 01

  Yeah. And so, um, yeah, you want to edit the output and you know, yeah, something like a notebook or AI Studio. There's not too many things that let you edit the output. But once you do that, once you like do that hard work of like that, those examples, especially like the thing you're trying to write now, then it starts to work really well.

• 52:44Claire Vo?

  Yeah, it was one of the most important things that I built into my, my AI product was every asset that gets generated has a real time editor for the user to update. And then those updates go back into the model because I just think if the central value proposition of your product is writing, which mine is, it's one of the hardest stylistic challenges I've seen AI struggle with. It all sounds like slop. Like I could identify AI writing from a mile away. And so, yeah, I found this like, um, incremental optimization, first outline, then draft, then edit, then refine, process takes a while. Um, there's some latency in the experience, but it ends up netting higher quality. And then just like use it as a draft, edit it, get the system, get the system to be better. So

• 53:29Claire Vo?

  that's really, really great feedback.

• 53:31Speaker 01

  This is for chat P R D?

• 53:32Claire Vo?

  This is for chat party. Yep.

• 53:34Speaker 01

  Very cool.

• 53:35Claire Vo?

  Yeah, you know, I have high standards for writing too. So it was important to me. Well, this was so great. Where can we find you and how can we be helpful?

• 53:44Speaker 01

  Yeah, hamil.dev is my website. You can also find me hamilhussain on Twitter. And yeah, I'm teaching a course on Maven, as you know, about evals that go into all these subjects and very deeply.

• 53:58Speaker 01

  But yeah, that's where to find me.

• 53:59Claire Vo?

  Great, yeah. And for our listeners that don't know, Lenny's List is on Maven, including a How I AI section that I think features your course. So you can check it out there. Thank you so much for the time. It was super educational, very practical. I'm going to take these tips right away and go improve my own product. Have a great day.

• 54:18Speaker 01

  Yeah, thank you for having me on.

• 54:20Claire Vo?

  Thanks so much for watching. If you enjoyed the 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.