Script to generate podcast titles, intro text, meta descriptions with OpenAI API (#74)

* Some scripts, topics for podcasts

* New timestamps for around 20 top podcasts (by topical popularity)

* Moved from folder

* Fixed some of the podcast issues

* Tests for intert_podcast_timestamps and fixes

* Script to generate podcast titles

* A script to generate intro for podcasts

* updates to scripts

* Updated titles, intros

* script for meta descriptions and updated pages

* Added update date

* Updates

Author: kavaivaleri

_podcast/s03e04-interviewing-300-data-scientists.md

@@ -1,11 +1,14 @@
 ---
-title: What I Learned After Interviewing 300 Data Scientists
-short: What I Learned After Interviewing 300 Data Scientists
+title: 'Data Science Interview Guide: CV Optimization, Take-Home Projects, Mock Interviews
+  & Negotiation'
+short: 'Data Science Interview Guide: CV Optimization, Take-Home Projects, Mock Interviews
+  & Negotiation'
 guests:
 - olegnovikov
 image: images/podcast/s03e04-interviewing-300-data-scientists.jpg
 season: 3
 episode: 4
+date: 2025-11-07
 ids:
   youtube: AYi7b-8GPm4
   anchor: What-I-Learned-After-Interviewing-300-Data-Scientists---Oleg-Novikov-e10ctbs
@@ -16,6 +19,13 @@ links:
   apple: https://podcasts.apple.com/us/podcast/what-i-learned-after-interviewing-300-data-scientists/id1541710331?i=1000520681105
 transcript:
 - header: Introduction & Episode Overview
+- line: This week we will talk about the interview process, getting hired as a data
+    scientist — and not only data scientists. We have a special guest today — Oleg.
+    Oleg worked as a data science manager at Uber, where he built data science teams.
+    He also has experience building several startups in Europe. Recently he created
+    NextRound which is a free service for practicing interviews, receiving personalized
+    feedback, and learning materials. Welcome!
+- header: Introduction & Episode Overview
 - line: This week we will talk about the interview process, getting hired as a data
     scientist — and not only data scientists. We have a special guest today — Oleg.
     Oleg worked as a data science manager at Uber, where he built data science teams.
@@ -917,6 +927,27 @@ transcript:
   sec: 4194
   time: '1:09:54'
   who: Alexey
+intro: How do you make your data science application stand out, ace take-home projects,
+  and negotiate an offer without leaving money on the table? In this episode, Oleg
+  Novikov — creator of NextRound and former data science manager at Uber with a background
+  in data and software engineering — walks through a practical data science interview
+  guide covering CV optimization, take-home projects, mock interviews, and negotiation.
+  <br><br> We dig into career trajectory from engineering to product data science,
+  building projects that differentiate your application, and concrete product work
+  like forecasting and LTV. Oleg demonstrates NextRound's mock-interview chatbot and
+  personalized feedback, explains common hiring funnels (recruiter screen → take-home
+  → interviews), and contrasts product data scientist vs. machine learning engineer
+  expectations. You'll hear specific advice on treating your CV as a landing page,
+  highlighting personal contributions, crafting case-study narratives from business
+  goals to evaluation metrics, and preparing for technical assessments (ML fundamentals,
+  SQL window functions, coding). We also cover handling rejection, replying graciously,
+  evaluating offers, negotiation tactics when your current salary is low, and practical
+  steps for PhDs breaking into industry. <br><br> Listen for actionable steps to refine
+  your data science resume, prioritize take-home ROI, and use mock interviews to iterate
+  faster.
+description: Master CV optimization, take-home projects and mock interviews to land
+  data science offers—learn SQL/ML prep, negotiation tactics and measurable project
+  impact.
 ---
 Links:
 

_podcast/s03e07-market-yourself.md

@@ -2,17 +2,25 @@
 episode: 5
 guests:
 - valeriybabushkin
-intro: In this episode, Valerii Babushkin—then Head of Data Science at Blockchain.com
-  and Kaggle Grandmaster—breaks down how to approach machine learning system design
-  at scale. He shares insights from building ML systems at Meta, Alibaba, and Yandex,
-  explaining how to move beyond algorithms to focus on end-to-end design, feature
-  engineering, and evaluation. Valerii walks through a real-world fraud detection
-  example, discusses how to structure interview answers, and outlines the core principles
-  from his book Machine Learning System Design. You’ll learn how to think like a senior
-  ML engineer and design robust, production-ready systems.
-description: Master ML system design interviews with Valerii Babushkin, ex-Meta Head
-  of Data Science. Learn fraud detection systems, feature engineering, metrics selection,
-  and production ML best practices for FAANG interviews.
+intro: 'How do you approach ML system design interviews that probe production constraints,
+  fraud detection trade-offs, and MLOps realities? In this episode, Valerii Babushkin
+  — Senior Director of Data, Analytics, and AI at BP, Kaggle Competitions Grandmaster,
+  and author of Machine Learning System Design — walks through what interviewers look
+  for and how candidates should structure answers for real-world ML problems. <br><br>
+  We cover concrete topics you can use in interviews and on the job: distinguishing
+  software vs. ML system design; a fraud detection case study (probabilities, loss
+  functions, real-time requirements); label noise, class imbalance, and feature engineering
+  trade-offs; end-to-end pipeline items like metrics, baselines, A/B testing, and
+  validating in production; monitoring, distribution shift, fallbacks, and production
+  robustness; serving models, embeddings, and MLOps roles; plus when to avoid ML and
+  practical checklist items for core projects. Valerii also shares interview tactics
+  — signposting depth, stating assumptions, iterative baselines — and guidance for
+  new grads and career progression toward system design roles. <br><br> Listen to
+  learn actionable frameworks, example trade-offs, and preparation strategies to improve
+  your ML system design interviews and production ML decisions.'
+description: 'Master ML system design: fraud detection, feature engineering & A/B
+  testing to ace interviews, build robust production models, monitoring and MLOps.'
+date: 2025-11-07
 topics:
 - machine learning
 - career growth
@@ -27,10 +35,13 @@ links:
   youtube: https://www.youtube.com/watch?v=0RsmRjar66E
 season: 7
 short: Machine Learning System Design Interview
-title: Machine Learning System Design & Interview Strategies for Senior ML Engineers
+title: 'ML System Design Interviews: Production ML, Fraud Detection, Features, A/B
+  Testing & MLOps'
 transcript:
 - header: Podcast Introduction & Episode Overview
 - header: 'Valerii Background: Career Snapshot and Kaggle Achievements'
+- header: Podcast Introduction & Episode Overview
+- header: 'Valerii Background: Career Snapshot and Kaggle Achievements'
 - line: This week, we'll talk about machine learning system design interviews. We
     have a special guest today, Valerii. Valerii works at Blockchain.com as a head
     of data science. Before that, he worked in quite a few places. More recently at
@@ -66,6 +77,7 @@ transcript:
   time: '3:06'
   who: Alexey
 - header: 'Blockchain.com Role: Scope, Responsibilities, and Data Ownership'
+- header: 'Blockchain.com Role: Scope, Responsibilities, and Data Ownership'
 - line: Well, sure. Let's start from the current time. As you said, I'm head of data
     science at Blockchain. So a bit about blockchain, first. It's a very old crypto
     company. When I say very old – it is very, very old. It was founded in 2011. Try
@@ -105,6 +117,7 @@ transcript:
   time: '5:42'
   who: Alexey
 - header: 'Transition to Meta: User Privacy Work and Large-Scale ML Experience'
+- header: 'Transition to Meta: User Privacy Work and Large-Scale ML Experience'
 - line: To some extent, yes, because it's everything related to data – from infrastructure
     to applications. From analytics to visualization. Before that, I was working in
     – well, I joined Facebook and left Meta. I will just rotate my screen a bit –
@@ -134,6 +147,7 @@ transcript:
   time: '7:30'
   who: Alexey
 - header: 'Hiring Experience: Conducting High-Volume Interviews and Team Leadership'
+- header: 'Hiring Experience: Conducting High-Volume Interviews and Team Leadership'
 - line: 'Live interview? Okay. I don''t think it''s about Blockchain’s mission. That''s
     it. What else? I was leading quite a big team in my time – the biggest team I
     was leading was almost 150 people: machine learning engineers, data analysts,
@@ -175,6 +189,7 @@ transcript:
   time: '9:07'
   who: Valerii
 - header: 'Candidate Targeting: Who Faces ML System Design Interviews'
+- header: 'Candidate Targeting: Who Faces ML System Design Interviews'
 - line: Okay. Let's talk about machine learning system design. This is a part of the
     interview process and you said you did a lot of interviews as the interviewer.
     I imagine also, when you were joining Facebook before that, you also had to take
@@ -217,6 +232,7 @@ transcript:
   time: '11:20'
   who: Alexey
 - header: 'Interview Structure: 45-Minute Narrative and Evaluation Goals'
+- header: 'Interview Structure: 45-Minute Narrative and Evaluation Goals'
 - line: Yeah, true. Good catch. Yes, level five is a Senior in terms of the level
     on Facebook, which means that, if you're on this level, it is an honorary thing
     to be on this level forever. So if you ended on level four, it was probably because
@@ -262,6 +278,9 @@ transcript:
   time: '13:36'
   who: Alexey
 - header: 'Contrast: Software System Design Versus ML System Design'
+- header: 'Fraud Detection Case Study: Probabilities, Loss Functions, and Real-Time
+    Needs'
+- header: 'Contrast: Software System Design Versus ML System Design'
 - header: 'Fraud Detection Case Study: Probabilities, Loss Functions, and Real-Time
     Needs'
 - line: Okay, let's try to determine the disparity between those two. First of all,
@@ -309,6 +328,7 @@ transcript:
   time: '13:58'
   who: Valerii
 - header: Labeling, Class Imbalance, and Feature Engineering Tradeoffs
+- header: Labeling, Class Imbalance, and Feature Engineering Tradeoffs
 - line: Fortunately, the very basic log loss is good here. So we know that we might
     start from log loss. We also know that we might start from a very basic linear
     regression model. Why is that? Because we know that it has to be very fast – in
@@ -375,6 +395,7 @@ transcript:
   time: '16:43'
   who: Valerii
 - header: 'Interview Tactics: Stating Assumptions and Getting Alignment'
+- header: 'Interview Tactics: Stating Assumptions and Getting Alignment'
 - line: That's quite a lot of information. I was trying to process this. That's quite
     a lot of things. So this was an example of machine learning system design. The
     interview starts and then the person – the interviewer – asks you, "Let's design
@@ -396,6 +417,7 @@ transcript:
   time: '21:10'
   who: Valerii
 - header: 'Example: Points-of-Interest System vs Personalized Recommender'
+- header: 'Example: Points-of-Interest System vs Personalized Recommender'
 - line: Yeah, indeed. So, the original question I actually asked you is about the
     difference between system design and machine learning system design and I think
     it's very clear what machine learning system design is. It requires some domain
@@ -459,6 +481,7 @@ transcript:
   time: '24:27'
   who: Valerii
 - header: 'End-to-End ML Pipeline: Metrics, Baselines, and A/B Testing'
+- header: 'End-to-End ML Pipeline: Metrics, Baselines, and A/B Testing'
 - line: But where does system design actually come into the picture here? Because
     here, we talked about selecting the right metric, which was the important thing,
     as you said. You said it was log loss for this specific case. Or even before log
@@ -558,6 +581,7 @@ transcript:
   time: '28:28'
   who: Alexey
 - header: 'Securing the Interview: Iterative Baselines and Signposting Depth'
+- header: 'Securing the Interview: Iterative Baselines and Signposting Depth'
 - line: Let's be honest, the interviewer was a human, and humans are subjective. Maybe
     they had a bad day. However, to some extent, I'm surprised because it's hard to
     say the interview was nodding. Maybe, again, the way you remember it and the way
@@ -602,6 +626,7 @@ transcript:
   time: '31:09'
   who: Alexey
 - header: 'Appropriate Depth: Practical ML Decisions vs Research-Level Detail'
+- header: 'Appropriate Depth: Practical ML Decisions vs Research-Level Detail'
 - line: Well, it's an interesting question for which there is no single answer. It
     depends. My opinion is that the interview has to be as close to the real job –
     the real work – as it can be. So, to be honest, in applied machine learning, you
@@ -640,6 +665,7 @@ transcript:
   time: '33:19'
   who: Valerii
 - header: 'Preparation Strategies: Mock Interviews, Resources, and Experience'
+- header: 'Preparation Strategies: Mock Interviews, Resources, and Experience'
 - line: Okay. [laughs] So, how do I actually prepare for machine learning system design
     interviews? It feels as though just being a practitioner is not enough. Because,
     first, you never know what exactly is expected. I guess you need to ask that.
@@ -726,6 +752,7 @@ transcript:
   time: '37:28'
   who: Valerii
 - header: 'Industry Checklist: Core ML Project Review Items and Patterns'
+- header: 'Industry Checklist: Core ML Project Review Items and Patterns'
 - line: Speaking of this mock interview – a while ago, I had a mock interview with
     Valerii, where Valerii interviewed me. The question was about designing a fraud
     detection system.
@@ -772,6 +799,7 @@ transcript:
   time: '39:13'
   who: Valerii
 - header: 'Defining Goals and Proxy Metrics: Business Alignment and Long-Term Health'
+- header: 'Defining Goals and Proxy Metrics: Business Alignment and Long-Term Health'
 - line: So about this checklist – let's say we need to design a system, not necessarily
     for an interview, but just design a system. What is the first thing we need to
     do? Do you remember what is in this checklist?
@@ -856,6 +884,7 @@ transcript:
   time: '44:01'
   who: Alexey
 - header: Features, Labels, Model Selection, and Validation Workflow
+- header: Features, Labels, Model Selection, and Validation Workflow
 - line: Let's say we know what we would like to do. We know how we can try to optimize
     it in this way. What does that mean? That means that if my model improves, there
     is a high chance that my metric of interest will be better. Now, I need to think
@@ -886,6 +915,7 @@ transcript:
   time: '44:11'
   who: Valerii
 - header: 'Production Robustness: Monitoring, Distribution Shift, and Fallbacks'
+- header: 'Production Robustness: Monitoring, Distribution Shift, and Fallbacks'
 - line: Perhaps if you cover all these parts during your system design interview,
     you're already in quite a good position. Right?
   sec: 2762
@@ -933,6 +963,7 @@ transcript:
   time: '47:48'
   who: Valerii
 - header: 'System Components: Why Features Matter More Than Model Architecture'
+- header: 'System Components: Why Features Matter More Than Model Architecture'
 - line: Okay. So let's go to the questions. We have quite a few of them. The first
     question we have is, “What are the typical components of a machine learning system?
     And what percentage of it are machine learning algorithms?”
@@ -987,6 +1018,7 @@ transcript:
   time: '49:57'
   who: Valerii
 - header: 'Engineering Integration: Serving Models, Embeddings, and MLOps Roles'
+- header: 'Engineering Integration: Serving Models, Embeddings, and MLOps Roles'
 - line: Thank you. Let's go to the next one, “How to make machine learning algorithms
     work with other parts of systems to solve real world problems?” I guess the question
     is more about, “Okay, we have this model that we just discussed. This model for
@@ -1020,6 +1052,7 @@ transcript:
   time: '52:14'
   who: Alexey
 - header: When to Avoid ML and Useful Design Pattern References
+- header: When to Avoid ML and Useful Design Pattern References
 - line: Do we really need machine learning here exactly? Maybe we can be lucky and
     we can just avoid it.
   sec: 3145
@@ -1072,6 +1105,7 @@ transcript:
   time: '53:59'
   who: Valerii
 - header: 'New Grad Expectations: Coding Focus and Limited System Design'
+- header: 'New Grad Expectations: Coding Focus and Limited System Design'
 - line: Yeah, so another question from Alvaro. Alvaro is graduating soon and he is
     a machine learning intern at a startup. He's starting a job hunt, hopefully [inaudible].
     So how much system design should he expect as a new grad?
@@ -1150,6 +1184,7 @@ transcript:
   time: '57:20'
   who: Valerii
 - header: 'Validating in Production: A/B Tests, Causality, and Human Labels'
+- header: 'Validating in Production: A/B Tests, Causality, and Human Labels'
 - line: Okay. I don't think we have a lot of time for more questions. There is an
     interesting question from Vijay, which is about, “What is the best way to validate
     the model performance in production? Do we need humans for that or are there other
@@ -1197,6 +1232,7 @@ transcript:
   time: '58:47'
   who: Valerii
 - header: 'Career Path: Moving from Data Science Practice to System Design'
+- header: 'Career Path: Moving from Data Science Practice to System Design'
 - line: Yeah, so the question is, “With this profile, you're very good at doing data
     science stuff. How did you transition from data science to being good at system
     design?”
@@ -1224,6 +1260,7 @@ transcript:
   time: '59:43'
   who: Valerii
 - header: Closing Remarks and Contact Information
+- header: Closing Remarks and Contact Information
 - line: '[laughs] Okay, I think that''s all we have time for. So maybe last one –
     How can people find you?'
   sec: 3603