Ace Analytics Interviews with AI

Sravya Madipalli: Awesome, awesome. Great. So let’s get started. Hopefully Zoom doesn’t do this to me again in this… Cool. So… Guess what? Let’s jump right in. Okay, give me a second. So here today, we are going to talk about acing analytics interviews with AI, right? So basically, we are going to talk through analytics interviews, but if you know, analytics interviews are pretty common for roles that are data science, product analytics, or regular data science, or data analyst, or even product. like, PMs, right? So I’m hoping, you are in one of these roles, majorly. Even if you’re not, whatever, details and tools we’ll discuss with AI is going to be apt for you, because we are going to talk through whatever is the domain that you’re going to work in, how can you use AI to just prep really well, for your interviews, and, you know, get.

a lot more access to different tools, and how do you access these tools to help you, like, be better prepared. Okay. So, today’s. like, lightning lesson is going to be based on one fictional character called Priya, and let’s say she’s an analyst, and she’s basically trying to get, herself prepared for these interviews, okay? Cool. So, before we go ahead and get into demos and, like, how to use different AI tools, I have one question, right? So… Let’s start with this question like almost everyone would get when you apply for a company. I took here Nextdoor as an example. If you know about me and Shane, we both worked at Nextdoor along with Hai, who’s not here today. So let’s say you’re now interviewing for Nextdoor. This is like the first question that you’d get asked by the recruiter, right? Why do you want to work at Nextdoor?

So, what does a generic answer there looks like? Which is not a bad answer, per se, right? For things like, I love your mission of building community, and I think my analytics background would be a great fit for the growth team. Not a bad answer. But, guess what, is an answer that’ll stick more, that’ll make sure that you definitely get, you know, a high score, or, like, a pass. For sure, or a strong yes, for a question like this. An answer would be something that’s make sure that you ground your answer in the experience. If it’s a technical question, then you ground it from your technical expertise in your company, but if it’s a behavioral question, like this one, then you ground it in your personal experience. So a grounded answer here could be when my neighbor’s dog went missing last spring, her next door post had a dozen people go out searching within an hour. To me, that’s your product’s best moment and it’s measurable. How fast a post gets a real response from neighbors and that’s the problem I want to work on. So what you’re doing here is you’re making your answer way more personal and getting it.

Like… To make sure that, like, getting it to a place where the recruiter or the interviewer actually remembers you and understands. like, why, like, what, like, way more about your answer, and that your answer’s not generic, but you actually invested in getting into the role. So you want to make sure you basically give answers that are grounded in your reality, in your experience, in your technical depth, in your, like, domain knowledge that you’ve gathered over all these years, right? For behavioral, it could be your own, like, personal, like, anecdotes like this one that I have here. For technical, it could be the type of work that you’ve done in this space before. Awesome. So, especially with AI, guys, the more and more people are preparing with AI, AI can write the first answer for anyone, right? The answer that’s generic. But regarding the answer that’s more grounded in your experience.

You need to basically give it your stories. You need to know your stories. You need to write down your stories. Both personal and also technical, for you to ground your answers from your experience. Awesome. So, why does, like, most prep go sideways? I generally see three reasons, right? So, the problem one, which is generic practice. So, I kind of gave you an example of what a generic practice and generic answers would look like in the prior question, right? So, basically, you ask for interview questions, you get textbook questions, and you end up practicing almost textbook answers, right? Nothing in them, in those answers is actually yours. That’s from your experience, your personal, like, domain expertise, or your technical. input that you’ve gathered over the years, right? So, that’s the first problem, like, pretty, generic answers, and what you do is, like, a generic practice, right?

And the second is that we read instead of talk. For example, like, it’s almost like, let’s say you’re doing mock interview with AI, right? Imagine reading your notes, or going through your cheat sheets, and, like. Almost by getting it by heart and, like, reading them through, but instead of… Using your, hand gestures, or in… you’re using, the words that look like you’re actually thinking out loud, rather than you’re… Sharing an answer that you’ve… by heart at all, most, and just repeating it. And it’s pretty evident, as someone who’s taken multiple interviews, like, hundreds of interviews over my career, you would know if an answer is prepared, rehearsed, to such an extent that It does not, like, feel… that you’re, you’re, like, speaking with, enough thought and, like, intent in that moment. So that’s a problem that I’ve seen in most of my interviews. And the other problem that I generally see is people don’t

prepare for specific interviews, right? For example, they generally have one mock for multiple rounds. They do not have specific, like, types of preparation for different styles of interviews. For example, a recruit… I’ve seen, when I also coached people, that people generally don’t prepare for a recruiter screen. They kind of go in saying that, oh, this is a recruiter, you kind of think you’re definitely going to go into the next round, and people are not prepared with really good answers for a recruiter screen. So the mock interviews, or the mock answers you prepare for Recruiter screen is going to be. totally different than your answers that you get prepared for a hiring manager screen, which are very similar, but they are looking at different perspectives of who you are and what you bring to the company that they are interviewing for, right? So these are, like, 3 types of, mistakes that I generally see. So…

Another thing that I also see, especially with AI these days, when I talk to multiple people that I try to coach, is that asking AI about, can you give me interview questions without giving all the context about your domain and the type of role you’re interviewing, and getting into the depth of the type of interview that you have scheduled in the next, like, say, next few weeks. He’s going to… just give you questions that are very generic, again. So… If I want you to take away something from this, is that make sure you have custom prep, that’s based out of, especially with AI, that is based out of the context of who you are, what your experience is, and what type of interview you’re going for, and what exact role that is for, rather than going for generic preparation and generic, like, questions and answers.

Okay, so… let’s say, what’s the first thing that I’m… that I’d suggest any of you that you start using AI to basically prepare. In fact, you don’t even, even in the world where there’s no AI, you probably need all of this written down. But with AI these days, it’s even more easier for you to get all of this together. So what am I talking about? I am here referring to you building your experience database before your first mock. What do I mean by this? I basically mean you working with, let’s say, you have so many note-takers these days, right? You could record, yourself in Granola, Otter.ai, Visperflow now has a note-taker. You basically, let’s say, even as simple as recording yourself in your, you know, iPhone voice recorder. I’m talking through your experience and questions related to multiple, like,

interviews that you need to, like, you know, prep for. For example, recruiter and hiring manager interviews. You need to talk through your behavioral aspects of your experience. So, who you are. Let’s say, introduction. Right? You know? Who you are, where you come from, what your domain has been, and… what are your, like, you know, leadership principles, if you’re a leader, or what are your ways of working, if you’re a, you know, IC? Like, having all these actually Like, you speak yourself into a recorder, or you get yourself into, like, one of these, like, note-takers, and get a transcript of all your experience from the behavioral aspect, right? Questions related to, you know, the most impactful outcome of a project, and having at least 3 to 5 impactful projects, and questions related to a friction with a stakeholder, having multiple examples of that. Like, think… and whatever I’m talking about right now, the experience database, it doesn’t have to be interview-ready. It could be something that you just speak with all these uhs and ums.

What, it doesn’t have to be, like, structured, because what you’re trying to build here is. like, a text transcript database that you can feed to AI to… so that you basically have your story cards ready, so that AI could help you structure your experience and all…

Shane Butler: Oh, I think we lost Sravya. Let me, send her, Message. She may not know she’s, frozen. And yes, we will, absolutely share out the recording. We’ll actually automatically send out the recording. Okay, looks like she dropped. She’s gonna come back in now. Sorry for the technical difficulty. So Sravya was talking about, like, building out your experience database. I think… She’ll get into, like, like, when you do this. I think there’s, like, a couple approaches when I’m kind of building out mine. One is, like, really setting some time aside. I know someone dropped in the comments here, like, you know, putting in the job description, your resume, etc. Like, you could sit down and really be thinking in the context of a single job or application and start building stuff out. I also find that there’s… I mean, there’s just, like, a lot of stories throughout all of our careers and all of our lives, that can be used. They don’t always come to us when we’re, like, sitting down for an hour or two, trying to get them all into a database. I take a lot of…

Voice memo notes. Like, when one just pops up into my mind. And then what I’ll do later on is I’ll… paste it into… Claude Code, or Codex, or whatever you’re using, or ChatGPT or something if I’m using chat, not, like, a terminal version of it. And I’ll update that with, like… you know, what I… what just happened to come to mind. at that, given point. What AI tool is best for this task? We need to make an agent for this. Yeah, you’ll probably, like, build out some agents for this. Sravya actually has a repo that she has created that she’ll share, later on for everyone to basically get started with. You can use Claude, you can use Codex, you could use… Like, the OpenStar 2, I probably want to use, like, Claude Code rather than chat, but we’ll see what she has. You could probably use Cowork, you could use…

ChatGPT work. We’ll definitely share that repo later on, and we’ll also send it out in an email later on, too.

Sravya Madipalli: Okay, but before that, guys, I’ll talk through what we’re going to get through today, so that, you know, you get excited about what’s coming. I’m going to share with you how you could basically use AI to, both web AI like cloud.ai and chatgpt.com and also cloud code. And I have a repo for you, a starter repo that will basically help you get started on… like, from the get-go. So there’s a… we basically have, I basically created, like, these, like, important skills. like Kickoff. So as soon as you use Kickoff, it’s going to basically ask you a bunch of questions about your experience, and you feed all your experience into it, so that you create a database. That you could work with and. Yeah, so you create a database that you could work with, that’ll basically help you do all the mock, interviews and, you know, all of that better.

Cool. So, let me actually share my… if you all can maybe pin… My… Screen, I’ll just share it from here.

Shane Butler: Get me loaded up. Yeah. Until now.

Sravya Madipalli: Yeah, until, I don’t know, internet, everything’s going crazy in my home today. I don’t know why. Cool. So… Yeah. So what basically, how do we prep with AI today? We are going to do basically something called kickoff. You talk through your experience, and AI builds the database, right? So you talk through your experience, multiple styles of experience. Experience that’s related to… that’s awesome, thanks so much, Shane.

Shane Butler: Which slide are we on here? We’re on,

Sravya Madipalli: I’ll be on side five.

Shane Butler: Kick off.

Sravya Madipalli: Yes. Okay. Cool. So, basically. you… we… we do a, we do a kickoff, and talk through your experience, and AI builds the database for you, right? And the second one, the second stage is. practice or simulate. So answer one question, or watch a model, and and then, you know, answer… all the answers are built from your own stories. So how do you practice and simulate based on these, you know, these stories that that you added into the context of your database, right? So basically, you work with AI and use your own experience that you recorded through the database and get the practice out of it. And then you basically now go to the third stage where you actually record the practice mocks, where you’re working with your friend or you’re doing a mock interview with someone where you…

Take all the mock interview related transcript, and you work with AI. and get a debrief of how you spoke, what your answers were, and where do you think you could do even better? What was your answer, and what would have been the ideal answer? Yes, the person you’re doing the mock interview with would probably help you, if you’re doing a mock interview with an expert in that field, but you could also take that transcript and work with AI to basically get, you know. Of the ideal answers and where you’ve gone wrong and where you could do better and things like that, right? Cool. So, now coming to, like. Like a topic. that we could do here, right? So, what are the four rounds? Here, today, I’m basically talking about two rounds. Sorry, four rounds. The first round is the screen, so I can kind of combine a bunch of things into, like, one round, so that it makes it easy for you to prep.

Recruiter screen and hiring manager screen, right? Because both those screens look for the fit, the level, and the scope of the role, and why are you interested in the role, right? Because… what I’ve seen, especially when I coach people in this, and also when I personally interviewed for a bunch of roles for myself and my team, people generally Actually, good candidates, they have great experience, but they end up kind of underselling themselves, or they end up not talking about the most important aspects of, you know. What their experience could entail. In some of them, they you know that they’re making up answers live, that they are actually not quoting things out of their experience, and they’re like thinking through at the moment and coming up with things that are probably not based out of their experience, right? So these are some things that I’ve seen where people fail at in these hiring manager and recruiter screens.

And coming up with the case studies. So, case studies, are basically things that I would say that are product case analytics interviews, and also interviews that, basically. ask a take-home case study as well, right? So, you get a take-home, and you probably are expected to come up with the structure and the decision, and, you know, what are some things that in case you end up in a situation that is presented in the case, how would you actually deal with it in real time if you are part of their company and an employee there, right? Most of the people that I see fail at these interviews are… they’re basically reciting a framework that they’ve learned. There’s so many frameworks out there, especially working with AI now. It, you know, gives you so many things that you could literally, by heart, and just

say it for the sake of it, but you’re not grounding that into the question and you’re not grounding that into the role that you’re interviewing for. And another very common thing that I see as a mistake is people don’t come up with recommendations at all. They don’t come up with recommendations now that they see this particular situation, they go through all of these. What is something that is recommended out of this particular case study and these are the instances? What is the recommendation? And what is the decision that the stakeholder is supposed to make? They basically miss out on all of these things. If you’re looking for an interview where people fail at mostly, right? And here, things change based on the kind of role that you’re interviewing for. If you’re interviewing for a data scientist versus data analyst versus a PM,

You will probably have different flavors of this, but at the end of the day, you’re expected to, you know, share a recommendation and share something that people could, you know, take this work and do something about it. A lot of people actually miss that out. And the other most important thing that I see people missing out on is regarding the technical work is they basically talk through the technical concept without actually grounding the concept when asked in the experience that they have. that they’ve gathered over the years. So they basically know what a P value is, but they can’t come up with an answer when asked about how would you explain this P value to a non-technical person, or how would you explain this to a product manager? So that’s something that people struggle a lot.

When it comes to you know the technical interviews and coming to the behavioral interview regarding influence, ownership, and the ambiguity that you deal with on a day-to-day basis. So these are some things that you would actually see get more and more important as you ascend the levels. If you are a junior person, you’re expected to do a little bit, like, be strong at these answers, but if you’re interviewing for a senior person, then especially leadership, staff roles, or senior staff roles, you’re expected to be really good at Like, giving examples of how you’ve done, in a situation where you had, like, you know, ambiguous outcomes and, you know, things like that. Awesome. So, Yeah, so going to the next, can you go to the next screen, Shane? Yeah. So, this is…

Shane Butler: I think I wanted to… wanted to say, Sravya, just, like. like, in this last slide here, I mean, I think what you’re seeing a lot is, like. one of the major separators is ingraining stuff in, like, your actual experience, like, being able to say the technical skills, but in your experience, being able to say why you want to be there, but in your actual experience, being able to say the cases, but in your actual experience, and I think a lot of that comes back to just, like. Humans in general, like. love stories, like, we love storytelling, like, for the whole existence of humanity, you know? It’s like, it’s why actors and actresses get paid so much freaking money when you, like, watch TV and all these things, like, we love stories, and so, like. Everything’s just gonna hit home so much harder if you can have a story around it, especially one that’s, like, true to you, or true to the company that you’re talking to, and not just, like.

reciting off the textbook answers, like Sravya said.

Sravya Madipalli: Awesome. Any other questions, by the way, that are being asked? I saw a question pop up about the prompts. So I’m going to share with all the prompts that you could use to basically work with AI to get this. For you guys. Any other questions, Shane?

Shane Butler: I think someone asked, like, what was… what’s, like, the best kind of tool, AI tool, in terms of, like… like, are you using, like, chat? Are you using co-work? Are you using code? Can you use across any of these? I don’t know if we’ll get into that later, but that was one of the ones that was earlier.

Sravya Madipalli: Yeah, we will get into that later, but I could also quickly answer that right now. you could literally use any of these. Codex, Claude Code, like, desktop versions, and the terminal, like, actually have a Git repo, so I’m going to share a Git repo with you guys later today, and you could use a repo, work in Terminal. So I personally use Terminal a bunch. for all of this as well. So any of these work really well. The only thing is that the more advanced the tool is, the more custom and complicated you could work with AI. That’s it. But I would say for 60 to 70% of the cases, even your, you know, web works well as well. Like for 50% of your, like. prep help, even web goes a long way. Yeah. So, Shane, before I move ahead, there’s a demo as well that we could share. Maybe you could clone the repo. It’s in AI Analyst Lab. You could clone the repo, and I’ll tell you what to do, how to run the demo.

Yeah. Sorry, guys, we are dealing with a technical difficulty here with my laptop, it’s weird. Shane’s currently filling in last minute to help with these things. Okay. Cool. So, in the screen, so let’s go ahead with, you know, what do I have, right here. So, in the screen, most people describe a smaller job than what they have. This is another thing. I kind of briefed this into, when I was talking about the previous interviews. For example, I’ve seen this a bunch, where. when people are asked, what do you do, or, like, what is… how do you describe yourself, in your… and what is your day-to-day, people come up with things like, I support the growth team analytics, I’ve helped, you know, with, pricing experiments, but they do not quote numbers, they do not quote impact.

And the more you code the impact, the more you code the exact specific things, and let’s say even in experiments, how many experiments you’ve done it, the more your answer is actually going to be like. something that sticks better and something that makes you like, you know, a stronger person to be hired for for that role. For example, instead of saying, I support the growth team with analytics and I’ve helped with a few of their pricing experiments, you could say something like, I’m the growth team’s analyst. I own the activation metrics. I designed the pricing test, and I moved annual plans from 9% to 14%. Here you’re talking about what is something that. you could… use when it comes to questions related to the type of projects you worked on and the type of domain expertise you have and what do you actually own.

So basically listen for support, helped with, was involved in, and you know we, when it was you, I know numbers. So that’s something that the AI coach flags when you work with AI and do mock interviews with AI. Saying that you want to ensure that you don’t come out as a person that was a support person, but actually someone that developed things, that partnered with and actually owned things and acted like an owner, rather than someone who was on the sidelines and was in a support role. Okay. So. Let’s, next slide, Shane. Awesome. So, talk through your experience. And you don’t, like, write it, basically. That’s what I’m here talking about, right? So, how do you actually build this database? So, I’m talking… I mentioned with you, right, that we need to build the database and, you know, all of that, but how do you actually build it?

There are four sessions. I just started with something that four sessions of about 30 minutes. But it depends on you and the extensive amount of experience that you gathered. You could definitely start with something like this, but you could also make it more detailed and more into the domain and expertise, and you could talk through every aspect. Role that you’ve held in your experience, but I would say start with something like You know? 30-minute sessions of 4 types of, four types of experience. The first one would be work in scope. What you own, the metrics, and who depends on you, and the tools. So I have, this particular example and some of the examples more catered towards an analyst or a data science product analytics person, but you could take this framework and use it for any role that you’re interviewing for, okay? So the session one that you record yourself, and you get the transcript out of, would be the work and scope.

And the second session would be the best projects, right? So, an analysis that changed the decision, a test that, an experiment that you designed and you, you know, came up with, something that’s complicated that you definitely want to share with the hiring manager and later on in the onset interviews. So ensure that you record that as well. And the session three would be the stakeholders and how you worked with stakeholders. What was the pushback? What was the friction? How did you reframe a request? How did you, you know, like… respectfully disagree with the other person. So that would be the session 3. And session 4 would be about mistakes and growth. For example, where did you go wrong? Where did you learn from your mistakes? And where did you basically acted like an owner, like, and all of that. So ensure that you record all of these things, and I would start with something like 30 minutes each. Use any note taker, or let the

You could also let the AI interview you. I have prompts in the repo for you that you could use and you could literally type it in or talk to it. To the AI based on, like, you know, your experience and things like that. Right? And for managers, I would basically say add another session about how did you lead a team, what are some experiences, both good and bad, from, you know, being a manager, and how did you improve performance of your team, and how do you make sure that you have superstars on your team. Oh, satisfied, like all of that stuff. Add another session from your experience for managers. Okay, move to the next, one. Shane. Okay, so here, I basically… Shane, did you get a chance to copy the demo one?

Shane Butler: Yes, I guess I got it.

Sravya Madipalli: That’s good.

Shane Butler: Do you want me to share my, do you want me to share my VS code? Okay.

Sravya Madipalli: Yeah. So there are three things in this demo. I’ll basically share all these prompts with you guys. So you could do that copy-paste in even Cloud.ai, Shane. You don’t have to do it in the VS Code, the first two. The first two are basically in chat. Chat bots, okay? So, you could also obviously do that in Terminal as well, but I just want to make sure that everyone, can, you know, use.

Shane Butler: Justin Clyde at AI.

Sravya Madipalli: Awesome. So go to the cloud.ai. So, you could… you could paste and see You could copy-paste the prompt. So this is something, guys, I’ve basically built with AI, a fictional person called Priya. She’s an analyst, and she basically has her experience in it, and You work with the Claude, like the Claude Chat or ChatGPT Chat, and help it ask you questions. To basically ensure your experience is covered in depth. So, Shane’s currently helping me with that. Thank you so much, Shane, for all these.

Shane Butler: Is this, is this, is this prompt number one in the repo? The live kickoff?

Sravya Madipalli: Yes, it should be. It should be.

Shane Butler: Okay, got it. I’ll paste it in the chat here in case people want to read it also. If you want to try it on your own.

Sravya Madipalli: Yeah, it should be a really big chat.

Shane Butler: Oh, it’s really, it’s really big. Maybe it won’t feel like that. I’ll break it into two, but I’ll paste it in here and I’ll put it in, I’ll put it in two parts.

Sravya Madipalli: While it runs, Shane, you could browse through what is written there. What is written there? Like, you could browse through the prompt itself. Go to Show More.

Shane Butler: Yeah. Here we go.

Sravya Madipalli: Awesome. So look at this, guys. This is like an example prompt, right? So it says, I’m preparing for interviews and want to build my experience database, interview me about my work, like a curious friend over coffee, not like a hiring panel, right? Because the reason why we keep the stakes low at experience database is because you don’t want to worry about structuring your experience, you don’t want to worry about anything here. All you’re trying to do is just grab All the… brain, you know, like, content that you have about all the roles you’ve held in your, in your past career, and to put that, like, to dump that into, like, a text and transcript that AI could do. So now, Shane, like, you could, you could go down and see what AI gave you. So, you could see the answer here, right?

Oh, that’s awesome, Attendee. I see that. What if you’ve already written down? That’s great. Even if you’ve already written down, you could paste that in and ask AI if it thinks whatever you’ve written down already. It could be a notes, like you know, you’ve written down your notes. You could like take pictures of that and send it to AI too, right, and ask it to grade it. Grade it on how strong or how solid or how thin your experience is. Experience story card is, and it is going to help you make sure that your stories from your experience are solid and have all the content it needs for interviews, right? So now look at what AI said based on the role that we gave. So which of those roles are you interviewing for? And if you had to explain, and if you had to explain, how would you explain?

So, Shane, why don’t you give the prompt basically that says, give me a minute. You could say, for analysts, Is the role?

Shane Butler: Okay.

Sravya Madipalli: Yeah.

Shane Butler: Analyst. I would explain to my 75-year-old neighbor that asked me my idea, like, I just look at computers.

Sravya Madipalli: Ha ha ha ha. So actually, Shane, you could go to the demo runbook as well, and you’d get the prompt, but I can read out to you so that if that’s easier, you could copy paste.

Shane Butler: Is it, where is it in here? Is it in this?

Sravya Madipalli: Yeah, in the repo.

Shane Butler: the.

Sravya Madipalli: Yeah, demo runbook, you could find it, but I could, you know… Maybe… Like, give you the exact detail as well.

Shane Butler: Yeah, if you wanna paste it in.

Sravya Madipalli: Yeah.

Shane Butler: You’re in Slack.

Sravya Madipalli: Unfortunately, my computer, my Wi-Fi is not working right now from probably it’s my Wi-Fi problem the whole time. Okay, you can say, you can say you’re an analyst, and… Just type analyst, and, you know? I’m going to… But but. And give like an example, which of these roles are for you? And if you had to explain what you actually do all day, who’s. Can you type something that’s, like, relevant to that? I’ll try to, yeah.

Shane Butler: I worked with… Prod dev teams to help them… Prioritize. Their work each week, I usually… Give them some sort of weekly update of our… we look at our metrics, and we try and figure out which ones are up or down, and then what were either the changes in the product or the user behavior that’s driving those. And then that gives the engineers and the product manager, designers some ideas of, like, what we should focus on, for this sprint. I also spend some time where people just have, like, ad hoc asks where they need to, like, understand, like, what our users are actually doing, and if it aligns to, like, stuff that they’re telling them in interviews or something. To see if that’s actually happening in the product.

Sravya Madipalli: Yeah.

Shane Butler: So, like, in that case, like, I used… Like, a voice transcriber, and so it’s gonna… you know, it’s like, that’s more like how naturally I would talk about something, right? So I think it’s helpful to do that rather than trying to, like, type out something here.

Sravya Madipalli: So look at this. Look at what it’s asking you. It basically is asking you even more detail into this. And I just checked, Shane, we only have 12 minutes. So what we could do is this, I’m going to share all of this with you all guys. You could use this to basically build your own database. If you don’t want to, like someone just said, if you don’t have a database already, you could use this prompt to come up with details of your database as well. Now, we could — do you have the prompt tool handy, Shane? We could quickly go to a simulate mode there. You could go to chatgpt.com, or you could just try here. Just. We’ll do a new chat.

Shane Butler: do…

Sravya Madipalli: If you could go to a new chat. So guys, the next thing we are basically trying out is how do you take an interview? context, whatever… so I gave it a bunch of interview context myself to help you all with the demo here. But you need to give your interview context, your experience database, and you work with AI to do the practice mocks. to do, like, to see what is… what does a great interview look like? You can simulate an interview with AI. So I have a prompt for that as well. So, you basically, use that prompt to maybe start a new chat, Shane. Definitely.

Shane Butler: And this is… this is the one where you extract information.

Sravya Madipalli: So.

Shane Butler: The.

Sravya Madipalli: The first one is extracting the database, coming up with your own database. The second one is, now that you have all your database ready, how do you practice with AI? So this…

Shane Butler: Is this the one that’s…

Sravya Madipalli: Yeah, this is…

Shane Butler: We were looking at a transcript.

Sravya Madipalli: Yeah. So, it basically uses a transcript to help gauge how your answers are. If you don’t have a transcript, it just gauges based on whatever your answers give. For example, let’s say you’re doing a mock interview for a recruiter, and you don’t mention the most important impact numbers from your project. Then it’s going to grade you saying, hey, you’re only going to get a 6x10 because you missed the, like, you know, 10% to wide impact on this metric in your answer. So for it to do that for you, you need to ensure that your database and transcripts are shared. But if you don’t have that, it’s going to basically grade you based on your generic answers. So, Shane, you could go and say, simulate a mock, simulate an interview, simulate a case.

Shane Butler: All right.

Sravya Madipalli: So, let’s see. So… It’s basically giving you like. The case, you just joined a food delivery app, how would you approach this? So, it basically is giving you… you could have this in the coach mode, you could have this in the practice mode, and you could have it in, like, multiple modes, right? So, this is how it would look like. If you basically copy-paste the prompt that I share with you and work with it. Okay? So this is the second thing that I’ll share as part of our email that I’ll send. And Shane, for the final one, let’s do a two minutes of the final demo as well, and we could go into questions and answers later. So I have a repo that’s in the AI Analyst Lab. Did you get a chance to open the repo, Shane?

Shane Butler: Yes, let me, let me share my screen over there.

Sravya Madipalli: That’s awesome. Yeah.

Shane Butler: I have too many, too many things open. Where are we at?

Sravya Madipalli: So, basically, if you all are aware, we have a free repo called AI Analyst. Free repo, in the same, like, you know, our AI Analyst Lab GitHub, I added another free repo for you guys, I’ll share the link with you in the email as well, where you could literally, clone the repo. And. You could work with the repo to get, like. to… to basically do all of this in the repo itself. So, Shane, you could just do slash kickoff in that, and… you would basically… literally do what you’ve done in that, like, Claude.ai that Shane shared, and it’s going to basically help you with. Sharing your experience. Yeah. If you do a slash kickoff. It should. So, it should basically help you with all the stories that you’ve gathered, and maybe, Shane, you could walk through the repo as well, while this happens, so that people, like, so the repo basically contains all the… you could do a command plus, or people, you could zoom in as well. It contains, skills related to… so kickoff is basically a skill that will help you.

Gather your experience, build your experience database. And I have examples for you, debriefs and examples and all of that for you so that you could, you know, just. Understand what type of. So, how do you practice? You could click on that, yeah, and just see. So this is a repo that’s a starter repo for you. This is a free repo that you could use, clone, and literally start using a repo like this for all your interview prep, right? And guess what? This is something that you could… this is, yes, based out of Claude, but you could also, like. use this into Codex and ask Codex to make it Codex ready as well. Yeah. So I’ll share the first demo prompt. How do you build your own database through working with Claude, web agents? And the second demo is how do you do practice and mock interviews yourself? And the third one that we just shared is a free repo that basically helps you extract your experience and work with you and practice with you. And I also have some practice stuff for SQL and the basic technical things.

But this is a starter repo that you could use and build on top of it based on the roles that you’re interviewing for. So Attendee, it’s not three repos. The first two are prompts. prompts you could use in chat agents. And the third one is a repo. It’s GitHub repo that you basically go to any of your Claude code. You could also go to even ChatGPT desktop version as well, like the apps, and also share this GitHub link, and it is going to help you with that too. Okay? Cool. So we have lost…

Shane Butler: She’ll send out a… she’ll send out the repo in an email right after this. Nice. Smile, you got it already. Perfect. Thank you for sharing that for us.

Sravya Madipalli: That’s awesome. Okay, sorry everyone for all the technical difficulties today. It was totally not something I expected. I was actually supposed to run this myself. Shane was busy today, so I’m glad Shane came through and we did this. Okay, so let’s take any questions.

Shane Butler: No worries. Yeah, what questions do we got? Let me scroll up through the… How much is this product? This is free. And then all the prompts, when you get the repo, they’re actually all in here. So there’s a directory right here. So those first ones, when Sravya was having me paste stuff into Claude.ai. I’m just opening them in here, so… She has everything… set up in here, so you could copy it and put it in chat, or you can use it. I mean, if you’re using the repo here, like, you’re probably using, Claude Code or Codex to access it.

Sravya Madipalli: Yeah. And quickly, Shane, while people please keep asking your any questions that you have, you could unmute and ask as well. While that happens, Shane, could you share what if people want to know more, how could they learn more from us in our courses? So could you maybe share that one slide while people come up with questions?

Shane Butler: Yeah. Yeah, we offer two courses.

Attendee: Yeah, that’s right. We have…

Shane Butler: I have one around called AI Analyst for Everyone. I got a slide up here, AI Analyst for Everyone. That’s about kind of like analytical thinking. How do you, you know, frame questions? How do you work with stakeholders? How do you do root cause analysis? How do you do causal analysis? And then, you know, we walk through, like, what are the frameworks for those, but then how do you… execute this technically with AI, rather than having to leverage

Sravya Madipalli: I think we’re not seeing the slides.

Shane Butler: Yeah, I’m trying to… here we go, here they are. And then the other one we have is called… Let me share this. Share. Agentic Analytics. build an AI analyst, and this is actually about building the system. So I think Attendee mentioned something earlier in the chat where he had built out, like, basically his own homegrown kind of tableau. So this is about building out, like. Agents and skills and functions and pipelines required, so you can have, like, a… an analyst that does, a lot of your analysis for you. We go through, like, how do you do AI evals to make sure it’s not hallucinating and working, how do you do Context engineering and engineering. Those are our two primary, kind of, flagship courses. AI analytics for everyone, I think, is starting in a few weeks.

We also have a, discount for both of these. We’ll put these in the… in the email, too. We had a hand up here. Yeah, Attendee.

Attendee: So, if you’re able to hear me?

Shane Butler: Yep.

Attendee: So I am in a, I’m not looking for any, you know, technical role kind of thing for my interview, but I’m just trying to prepare myself for some kind of delivery role, which is a kind of generic development role. And I tried using, you know, like Gemini saying, hey, Google, prep me for an interview for this role. And it tried to answer this. But I didn’t find it very useful. Because every time you stop for a few seconds and then it goes, saying that, okay, you are gone. Something like that. So, I don’t know, whether I have to trust it. Whether to use some other tool, I don’t really know. So, what is the starting point? I mean, maybe I’m asking you a very basic question.

Sravya Madipalli: So, Attendee, you know, great question about, roles that are not… I’m talking about here specifically, right? Analytics or something. To be honest, you could use AI for any of these, roles, right? So, the most important thing For any of the interview prep, for it to get the context. So for it to know that who you are, what your experience is, which roles are you interviewing for, what did you interview in the past? What are the, you know, where are… Did you pass them? Did you… did not? What, according to you, were something that you did not do well at? What were your strengths? What did recruiter maybe share with you about some of your strengths, or, you know, weaknesses, or some… where you actually gone wrong? So, it needs all of this information for it to help you with the prep.

So yes, you could work with Gemini, Claude, ChatGPT in one-off chat sessions. But it’s going to be very relevant for that moment, like Attendee that you were saying. You chatted in that moment, and then let’s say you unlock or you basically have another chat, then you lose all the context. You have to work really hard to put all the context in that one chat session. That’s the reason I would suggest using a repo or using a memory that you basically save all of your The who you are, what you do, and what what are you interviewing for, and every detail of. Your interview journey so that it can help you prepare well from then on.

Attendee: I got it, I got it. I got a gist in bits and pieces from what you were explaining earlier, preparing the, you know, repo, etc. But, I’ll have a look through all the things which is coming through from you after the session.

Sravya Madipalli: Yeah, look at the email. I’ll try to answer it a little bit, your question through the email as well. And maybe you could get the recording again and, you know, listen through the recording. I know we had a bunch of glitches, but hopefully that will help you. Awesome. Yeah, Attendee, we could also do this in Gemini instead of Claude. What I’ve seen is Gemini is slightly not as grounded as Claude. It basically is more, I would say, user-like. It tries to make you happy as a person. Claude, on the other hand, is a lot more grounded, even the latest Codex models as well. You have to work pretty hard for Gemini to be grounded. You have to really add it in the prompts and other things to make it more grounded. Awesome. So someone was asking if there is an interview course, right? There would be an interview course. So what we could say, we are thinking of something like that.

To help you with, you know, how do you work with any interviews, with AI. So we’ll probably come up with something like that, so look out for that. Until then, though, in the AI analytics for everyone, we have a one week, a bonus week that gives you entire, like, set of like, database that you would need to interview for top 6 or 7 companies that have different domains. So we have that, we call it a bonus week, that’ll help you with interviews. So AI Analytics for Everyone is the course that you would Definitely be benefit tick, like, you know, it’ll be beneficial for you if you’re interviewing for any analytics roles because it. tells you about how do you master analytical thinking, how do you come up with metrics, how do you come up with, like, you know, causal experimentation, how do you tell good stories? So that’s literally how do you be a better analyst or a product manager analytics or data science product analytics person, which definitely will help you through interview as well, along with giving you a bonus week on interviews.

Yeah, so we have all of that. As part of analytics for everyone.

Shane Butler: Yeah, we didn’t… yeah, we didn’t touch too much on, like, how interviews are changing and stuff right now for, like, data analysts and data scientists, but as more companies are expecting folks to use AI or in their analytics workflow. I think, like, both of our courses really support that. We’ve had people take both of our courses. Just because they were in between jobs, and they’re interviewing, and they wanted to, like, kind of stay up to date. And we’ve had multiple people who are land, like, senior manager, data science, data analytics, positions. From… from basically citing the information they… they learn in those courses, but… We’ll come up with some interview stuff and and lots of more free interview stuff as well in the coming months. So keep an eye out.

Sravya Madipalli: Awesome. Cool. Once again, sorry guys for all the technical glitches. We’ll have these sessions again and again. That’s something that Shane, Hai, and I are thinking, to basically, you know, like, innovate, based on the newer models. How do we work with AI to basically get even better, at prepping for interviews and stuff? So look out for more lightning lessons from us as well, that we’ll probably have in, like. Two, three, maybe three weeks from now or something, we’ll try to have this as a cadence so that, you know, you all could learn more.

Shane Butler: Yeah, and you can find all of our upcoming Lightning Lessons At this link in the chat. So these are our free Wednesday… we host these every Wednesday at this time for free. So, if you want to sign up for any of the upcoming ones, we have a cool one next week on metrics. we have the one after that on… on building semantic layers. That one’s gonna be really cool, that’s, I think, the… Understanding semantic layers and context in that way is important. Right now, for landing any sort of data role. But yeah, that’s Attendee, I said we’ll add more interview-focused ones as well in the… coming weeks. Cool. Well, thanks everyone for joining. Thanks, Sravya.

Sravya Madipalli: Thank you guys. I’ll send an email.

Shane Butler: the.

Sravya Madipalli: Yeah.

Shane Butler: See ya.

Attendee: Thank you.