Data Neighbor Podcast

Ever wondered how someone with a political science degree ends up doing machine learning at Twitch? Meet Marina Wyss - applied scientist, blogger, YouTuber, and all-around productivity and learning expert. In this episode of the Data Neighbor Podcast, Marina shares her unconventional journey into tech, how she self-taught herself machine learning, and why her mantra of being “gratitude-driven” is her antidote to hustle culture.We dive into:- How Marina transitioned from political science and jewelry management into data science and ML.- Her self-study roadmap: From free courses to Coursera to deep technical books.- Practical frameworks for self-learning, getting promotions, and breaking into the ML industry.- Why she rejects hustle culture in favor of a gratitude-driven approach to productivity.- How she leverages AI tools like ChatGPT and Replit to accelerate learning and personal projects.- Common mistakes early learners make and how to avoid being overwhelmed.- Her take on Python vs R, what to focus on when starting ML, and why building your own projects is essential.Whether you're coming from a non-technical background, looking to break into ML, or trying to navigate learning in the age of AI, Marina’s story and advice will inspire you to take the leap - and build the skills that matter.

Links Mentioned in the EpisodeMarina’s Blog: https://www.gratitudedriven.com/Books Mentioned:- Designing Machine Learning Systems by Chip Huyen- AI Engineering by Chip Huyen- Software Engineering for Data Scientists by Catherine NelsonCourses:- Machine Learning Specialization by Andrew Ng (Coursera)- Deep Learning Specialization (deeplearning.ai)- Math for Machine Learning (Three Blue One Brown on YouTube)

Connect with Hai, Sravya, and Shane (let us know which platform sent you!):Hai: https://www.linkedin.com/in/hai-guan-6b58a7a/Sravya: https://www.linkedin.com/in/sravyamadipalli/Shane: https://www.linkedin.com/in/shaneausleybutler/#machinelearning #datascience #careertransition #gratitudedriven #selfstudy #ai #deeplearning #python #coursera #careerroadmap #productivity #chatgpt #learnML #dataeducation #DataNeighborPodcast #nontechtotext #womenintech #mlprojects

AI systems are becoming integral to nearly every digital product, but their vulnerabilities pose real and serious risks. How can companies protect their AI-powered products from security threats like prompt injection, jailbreaking, and misalignment?In this episode of the Data Neighbor Podcast, we're joined by Sander Schulhoff, CEO of Hacker Prompt and founder of Learn Prompting, to uncover critical security flaws in AI systems and practical ways to defend against them. With insights gathered from over 600,000 real-world AI exploits, Sander breaks down the three most dangerous AI security failures threatening today's products and provides actionable strategies to safeguard your systems.Connect with Sander Schulhoff:LinkedIn: https://www.linkedin.com/in/sander-schulhoff/AI Red Teaming Masterclass: https://maven.com/learn-prompting-company/ai-red-teaming-and-ai-safety-masterclassHack A Prompt: https://www.hackaprompt.com/Learn Prompting: https://learnprompting.org/Connect with Shane, Sravya, and Hai (let us know YouTube sent you!):Shane Butler: https://linkedin.openinapp.co/b02feSravya Madipalli: https://linkedin.openinapp.co/9be8cHai Guan: https://linkedin.openinapp.co/4qi1rYou'll learn essential techniques for securing AI systems, including how to recognize and prevent prompt injection and jailbreaking attacks, strategies for detecting misalignment early, and how to effectively leverage automated red teaming alongside human expertise. Sander also explores why security considerations must move from late-stage fixes to foundational aspects of AI model development and deployment.We discuss emerging security threats with autonomous agents, the role of government and compliance in AI security, and practical advice for teams at any stage—from startups to large enterprises—to proactively address AI security.If you're a data scientist, product leader, security professional, or executive interested in deploying secure AI systems, this episode provides critical insights and practical steps to protect your products and your users.

#AIsecurity #promptinjection #jailbreaking #redteaming #aisafety #machinelearningsecurity #aiattacks #dataprotection #aivulnerabilities #automatedredteaming #agenticAI #hackaprompt #aiethics #dataneighbor

Generative AI applications are transforming industries, but taking a GenAI model from prototype to production can be challenging. How can teams effectively build, evaluate, and deploy powerful generative AI systems in real-world scenarios? In this episode of the Data Neighbor Podcast, we're joined by Surabhi Bhargava, a Machine Learning Tech Lead at Adobe, to explore the step-by-step process of creating and productionizing GenAI apps, including embedding strategies, chunking techniques, retrieval-augmented generation (RAG), prompt engineering, and advanced model evaluation.In this episode, you'll learn essential insights into how to build a GenAI app, including how to select the right embeddings and chunk size, effective vector database management, and methods for robust query reformulation. Discover best practices for integrating GPT, Claude, Azure AI, and OpenAI APIs into your machine learning pipelines. Surabhi also shares critical tips on optimizing your AI prototype for user testing, identifying the ideal tech stack, and managing iterative feedback.We explore how to properly evaluate GenAI models using automated and human-in-the-loop strategies, discuss practical metrics to measure AI accuracy and performance, and reveal common pitfalls in AI application development. You'll also gain insights into personalization, user experience considerations, resource management, and understanding when not to use LLMs.If you’re a data scientist, engineer, product manager, or executive looking to deepen your understanding of generative AI and effectively move AI projects from concept to production, this episode is your ultimate guide.

Connect with Surabhi Bhargava:LinkedIn: https://www.linkedin.com/in/surabhibhargava/Connect with Shane, Sravya, and Hai (let us know which platform sent you!):Shane Butler: https://linkedin.openinapp.co/b02feSravya Madipalli: https://linkedin.openinapp.co/9be8cHai Guan: https://linkedin.openinapp.co/4qi1r

From Data Scientist to Product Manager to Software Engineer - all within Airbnb. In this episode of the Data Neighbor Podcast, we dive deep into Robert Chang’s remarkable trifecta career journey. If you've ever wondered how to pivot roles in tech, build social capital, or thrive across vastly different domains, this conversation is a must-watch.Robert shares:- His full-circle 10-year journey through data, product, and engineering.- Behind-the-scenes insights on building and scaling Airbnb’s metrics infrastructure.- The chaos and lessons from jumping into a PM role during a system-wide incident.- Why closing the loop and building with impact matters more than sticking to rigid job descriptions.- The real differences between maker vs. manager schedules - and how to find your fit.- The unexpected power of cross-functional empathy in building better teams and products.- How to embed learning into your actual work so it sticks and scales.- And of course… what “Hard Mode” really feels like (and why he chooses it, over and over again).Whether you're early in your career, thinking of switching roles, or simply want a raw, thoughtful perspective on navigating tech from within, Robert’s story will resonate. He doesn’t just talk about pivoting - he’s done it, with clarity, humility, and strategic intent.

Connect with Hai, Sravya, and Shane (let us know YouTube sent you!):Hai: https://www.linkedin.com/in/hai-guan-6b58a7a/Sravya: https://www.linkedin.com/in/sravyamadipalli/Shane: https://www.linkedin.com/in/shaneausleybutler/#DataNeighborPodcast #RobertChang #AirbnbTech #CareerPivot #TechCareers #DataScience #ProductManagement #SoftwareEngineering #InfraAtScale #MetricsPlatform #TechTransitions #PMLife #DataInfra #CareerGrowthInTech #CrossFunctionalSkills #ClosingTheLoop #GenAI #HardMode

AI is everywhere, but Responsible AI and AI ethics have never mattered more. How can teams ensure they build trustworthy AI systems free of dangerous AI bias and problematic algorithmic bias? In this episode of the Data Neighbor Podcast, we're joined by Will Landecker, founder of Accountable Algorithm, to dive deep into AI safety, ethical AI, and best practices for responsible development.
This episode explores foundational topics, including an introduction to Responsible AI, core Responsible AI principles, and practical strategies for addressing LLM hallucination, evaluating LLMs, and managing human-AI interaction. Learn what AI transparency, AI accountability, and AI fairness really mean, and how they impact your business and users.Will shares actionable insights on bias mitigation in AI, building fair AI models, and achieving AI explainability. We'll uncover the hidden AI blind spots that damage trust, discuss essential AI governance methods, and reveal tactics for effective AI risk management. From ethical machine learning frameworks to the latest strategies in detecting and preventing AI discrimination, this episode covers everything you need for developing truly ethical products.If you're a data scientist, product manager, engineer, or executive interested in the risks and rewards of AI, this is your essential guide to the complex intersection of AI and society. Discover how responsible practices aren't just good ethics—they're good business.

Connect with Will Landecker:Website: https://accountablealgorithm.com/LinkedIn: https://www.linkedin.com/in/will-landecker-88b2a788/Connect with Shane, Sravya, and Hai (let us know which platform sent you!):Shane Butler: https://linkedin.openinapp.co/b02feSravya Madipalli: https://linkedin.openinapp.co/9be8cHai Guan: https://linkedin.openinapp.co/4qi1r

#ai #responsibleai #aibias #aigovernance #aiethics #ethicalai #algorithmicbias #llmhallucination #aifairness #aisafety #aitransparency #aiexplainability #ethicalmachinelearning #biasmitigation #aiaccountability #aiandsociety #humanaiinteraction #trustworthyai #aiblindspots #risksai #machinelearning #aialignment #dataneighbor

Ever wondered why you might know more math, physics, or logic than you realize - and why traditional education often makes learning feel way harder than it should? In this mind-expanding episode, we sit down with Luis Serrano, founder of Serrano Academy to break down how real learning happens, why AI won’t replace the most human parts of us, and how quantum computing might just change the future of everything.In this conversation, Luis shares: - How struggling with math as a kid led him to a PhD and global teaching success. - Why "playing" is the best way to learn technical subjects like machine learning - not memorizing formulas. - The true power of intuition, emotional intelligence, and creativity in an AI-dominated world. - Quantum computing explained simply: why qubits, entanglement, and "magic" matter. - Actionable advice on how to learn faster, get past fear of complexity, and build irreplaceable skills for the future.Whether you're starting your journey in machine learning, curious about quantum computing, or wondering how to stay relevant in an AI-powered world, this episode is packed with practical inspiration and futuristic insights you don't want to miss. Watch till the end - trust us, it’s worth it!
Serrano Academy: https://www.youtube.com/@SerranoAcademyConnect with Hai, Sravya, and Shane (let us know which platform sent you!): Hai: https://www.linkedin.com/in/hai-guan-6b58a7a/ Sravya: https://www.linkedin.com/in/sravyamadipalli/ Shane: https://www.linkedin.com/in/shaneausleybutler/

What does it really take to be the first data hire at a startup?In this episode of the Data Neighbor Podcast, we dive deep with Jessie Li - former Head of Data at Descript and ex-Meta product data science leader - as she unpacks her raw, insightful journey navigating the unpredictable world of early-stage startups. From leaving Big Tech for a less-known startup to building a full-stack data function from scratch, Jessie shares the highs, lows, and the very real lessons learned along the way.We cover: - What shocked Jessie most on Day 1 as Descript’s first data hire. - Why bringing Big Tech best practices into a startup doesn’t always work. - How to build data foundations, insights functions, and culture from zero. - The push/pull model for career transitions (and how to think two jobs ahead). - Why being “too humble” in a startup can backfire - and how to avoid disorientation. - What startups really need in data hires - and how to know if you're a good fit. - The underrated importance of AI in data workflows today.Whether you're contemplating a startup move or want an honest look at what it takes to own and scale data in a high-growth environment, this is an unfiltered and inspiring conversation packed with lessons for data professionals at every level.

Connect with Jessie: https://www.linkedin.com/in/jessieljy/Connect with Hai, Sravya, and Shane (let us know YouTube sent you!):Hai: https://www.linkedin.com/in/hai-guan-6b58a7a/Sravya: https://www.linkedin.com/in/sravyamadipalli/Shane: https://www.linkedin.com/in/shaneausleybutler/#DataScience #Startups #JessieLi #FirstDataHire #Descript #DataCulture #DataEngineering #AIInData #CareerGrowth #BigTechVsStartup #DataNeighborPodcast #DataLeadership #IntentionalLearning #StartupHiring #WomenInData #DataDriven

In this episode of Data Neighbor, we dive deep into the rapidly evolving landscape of education driven by AI. Jess Haberman, Director of Product Content at Anaconda and former O'Reilly editor, shares groundbreaking insights on how generative AI tools are revolutionizing the way we learn Python, data science, data engineering, and beyond - making learning faster and more accessible than ever before.​ AI is even teaching us how to speak with ai, how to build ai, and how to learn how ai works.Key Takeaways:-How AI enables learners to master Python and technical skills significantly faster.-The future of traditional education, online courses, and tech bootcamps.-Critical soft skills you need to thrive in the AI-powered job market.-Why AI-assisted learning is now a necessity, not just an advantage.-Practical strategies to leverage generative AI for your professional development.​Connect with Jess:LinkedIn: https://www.linkedin.com/in/colinmatthews-pm/X: https://x.com/JessHabermanBluesky: https://bsky.app/profile/jesshaberman.bsky.socialWebsite: https://www.jesslhaberman.com/Connect with Hai, Sravya, and Shane (let us know which platform sent you!):Shane Butler: https://linkedin.openinapp.co/b02feHai Guan: https://linkedin.openinapp.co/4qi1rSravya Madipalli: https://linkedin.openinapp.co/9be8c

Ever wondered who’s behind the ads that actually make you buy stuff—and how companies waste millions on the ones that don’t? In this mind-opening episode of the Data Neighbor Podcast, we dive into the world of marketing science with Agastya Komarraju, Global Head of Product and Science for Growth & Funnel at Amazon. Agastya has held leadership roles at Walmart and Nielsen, and he’s here to demystify how data powers smarter marketing—and how Amazon slashed 90% of marketing spend without hurting results.- What exactly is marketing science and how does it differ from traditional data science? - Why “50% of your marketing spend is wasted”—and how to figure out which half. - How direct mail still outperforms digital ads (yes, seriously). - The five pillars of marketing science—from experimentation to attribution. - How AI and GenAI are transforming marketing from creative to targeting. - Real-world wins: The genius experiments that won Agastya and his teams top industry awards.Whether you're a marketing leader, data scientist, or just someone who wants to understand why you keep seeing that ad for a coffee mug you Googled once, this episode is packed with insights on consumer behavior, optimization, and the science behind persuasion. Learn how the biggest brands in the world are applying data—and what that means for your role, your budget, and your career.

Connect with Hai, Sravya, and Shane (let us know which platform sent you!):Agastya: https://www.linkedin.com/in/agastya-kumar-komarraju-95b60446/Hai: https://www.linkedin.com/in/hai-guan-6b58a7a/Sravya: https://www.linkedin.com/in/sravyamadipalli/Shane: https://www.linkedin.com/in/shaneausleybutler/

How is AI quietly reshaping your everyday workflow—whether you notice it or not? In this episode of the Data Neighbor Podcast, we explore how AI is being embedded directly into the tools you already use, with real-world examples from Adobe Acrobat.Our guest, Nikhil Pentapalli, is a Senior Machine Learning Engineer at Adobe, where he works on embedding GenAI into Acrobat to transform how people interact with documents. From AI assistants that can summarize and answer questions across multiple PDFs, to intelligent restructuring of scanned documents for mobile, we unpack what it really looks like to bring AI into production—and why it’s harder than it seems.We cover:- How AI is becoming invisible infrastructure in everyday workflows. - Why integrating AI into documents is harder than you'd think. - What makes PDFs complex, and how GenAI is making them smarter. - Building GenAI products: from prompt engineering to memory optimization. - Advice for breaking into GenAI, and what ML engineers need to know now. - Traditional ML vs LLMs: where to use which, and how to manage costs.Whether you're building AI tools, thinking about career moves, or just curious about what it takes to ship real GenAI products, this episode is loaded with insights and practical advice.

Connect with Hai, Sravya, and Shane:- Hai: https://www.linkedin.com/in/hai-guan-6b58a7a/- Sravya: https://www.linkedin.com/in/sravyamadipalli/- Shane: https://www.linkedin.com/in/shaneausleybutler/#genai #aiworkflow #documentai #rag #retrievalaugmentedgeneration #acrobatai #llm #promptengineering #mlengineer #aiintegration #realworldai #aiproducts #datascience #dataneighborpodcast