The Test Set by Posit

Vincent Warmerdam has been the first full-time hire at a startup, a spacey punster who accidentally got himself a job, a bartender at an Amsterdam comedy theater, and a Dutch bike tour guide — and he'll tell you all of it was career development. Now doing DevRel at Marimo, Vincent makes the case for reactive notebooks, Lego-brick widgets, and why "number go up" is not a data science strategy. Also: chickens die. The model doesn't know. This matters more than you think.What's insideHow a spacey pun accidentally launched Vincent's careerWhy Marimo's constraints make it better for LLMs, not just humansThe gorilla hiding in your dataset — and why the model missed itVibe coding vs. notebooks: three cells at a time as a disciplineWidgets as Lego bricks: reusable, composable, criminally underusedCognitive debt, confirmation bias, and sycophantic data scienceWhy natural intelligence is still, actually, a pretty good idea

Benn Stancil built Mode Analytics, spent a decade in the data trenches, and now writes some of the sharpest, funniest essays in the data world. On The Test Set, he talks about the cultural shift from Nate Silver to Rick Rubin why AI might kill the analytics dashboard, and what happens when a thousand startups all build the same thing. Plus: boy bands as a model for collaboration, and why the best creative work starts with cheating.What's inside: Why the modern data stack was basically big data 2.0The cultural flip from Nate Silver to Rick RubinGas Town, tar pits, and the AI startup zero-sum gameSoftware is becoming content, and that changes thingsBenn's creative process: Lorde lyrics, Codenames, and cheatingThe boy band as a model for small-team collaborationBI is (mostly) dead, and vibes might replace SQL

dbt Labs CEO Tristan Handy drops into The Test Set to map the fault lines between the data science world and the enterprise data world — and explain why analytics engineers are basically pissed-off data analysts who decided to organize the bookshelf. We get into SQL's glow-up, the community magic of dbt Slack, what AI agents mean for data warehouses, and why everyone's building iOS apps with Claude now.What's inside:What analytics engineers *actually* doSQL's journey from deeply unsexy to indispensableHow dbt turned source control into a source of truthBuilding a tech community without the RTFM energyAI agents on your data lake: permissions get personalWill LLMs kill the open-source package ecosystem?Edible gardening, welding dreams, and digital dysphoria

Nell Thomas has spent two decades in data — from equity research to the DNC to Facebook to leading a 400-person data org at Shopify. She walks Michael and Wes through the modern data stack role by role, gets honest about what AI is and isn't changing about data work, and admits the semantic layer has been her greatest leadership failure. Plus: Sneakers gets the respect it deserves.Episode NotesWhat does it actually look like to run data infrastructure for millions of merchants while the entire industry reinvents itself in real time? Nell Thomas (VP of Data, Shopify) talks vibe-coded dashboards, political campaign data scarcity, blameless postmortems, and why no one should be locking in on an AI strategy just yet. Recorded live in Times Square.What’s InsideMapping the modern data stack, role by roleWhy data quality is still the #1 problemWhat "good scrutiny" looks like on a data teamVibe coded dashboards and the trust problemShopify's MCP for their data warehouseThe throwaway tech problem in political campaignsWhy the semantic layer is so damn hardSneakers!

Phillip Cloud has been shaping the Python data ecosystem since the early pandas days — and he has *opinions*. Now a principal engineer at NVIDIA leading the Ibis project, Phillip talks about how he stumbled into open source via an eye movement lab, why he prefers his coding agents cold and emotionless, and what happens when you ask an LLM for woodworking trig. Plus: terminal user interfaces, the file hierarchy standard hot take nobody asked for, and the pineapple-on-pizza hill he's willing to die on.Episode NotesPhillip Cloud (NVIDIA, Ibis project) joins Michael Chow, Wes McKinney, and Hadley Wickham to talk about his path from eye movement labs to pandas core team, why developer productivity tools have quietly gotten amazing, his brutally honest take on coding agents, and what it would actually take to impress him. Also: VisiData love, NixOS evangelism, and yard work as therapy.What’s InsideFrom eye movement labs and MATLAB to pandas core teamColumn multi-indexes: the feature nobody likes but Phillip neededWhy your command line tool better have a sweet TUIVisiData: the terminal data tool you're sleeping onAre we writing code for humans or for agents now?Phillip's AI skepticism journey: Cursor, Claude Code, and frustrationThe Numba CUDA test suite port that would finally impress him

Charlie Marsh built Ruff, uv, and Ty — the tools that mass-fixed Python's worst pain points. Now he's grappling with what happens when agents start writing most of the code. In this episode, Charlie gets real about his team trusting his PRs less, the gnarly middle of coding with agents, and whether Python is even the right language for an agentic future. It's honest, a wee existential, and deeply relatable if you ship code for a living.Episode NotesCharlie Marsh is the founder and CEO of Astral — the company behind Ruff, uv, and Ty. He sits down with Michael and Wes to talk about what it's actually like building with coding agents every day, why his team's code review dynamics completely changed, and big open questions about code quality, open source community, and Python's future nobody has answers to yet.What’s InsideHow Ruff convinced mass adoption when switching tools is painfulWhy "just uv run it" became the killer featureHiring outside Python's ecosystem to build tools for itHis team said "we trust your PRs less now"Engineers screen-sharing their actual agent workflows at AstralThe "Lisp Curse" reborn: cheap code breaking open sourceIs Python the wrong language for an agentic world?

Alenka Frim went from teaching yoga full-time to becoming a committer and PMC Member on Apache Arrow. In this episode, Alenka joins The Test Set hosts to talk about how Arrow grew from spec to critical infrastructure, and why she started contributing to a project she had never even used. She reflects on imposter syndrome, the discipline of showing up (on the mat and in GitHub), and how agents are changing what it means to write code. Plus: managing 4,000 open issues without losing your mind.Episode NotesAlenka's path into Arrow is unconventional: Sshe wasn't looking for a job, she wasn't using the tool, and she'd spent the previous five years focusing on mind-body fitness. But open source felt like the right place to learn, have fun, and figure things out, so she jumped in. What followed was a journey from her first R bindings to becoming a PMC member on one of the most critical pieces of data infrastructure in the world.What’s InsideAlenka's journey, from yogi to Arrow committerSigns of a healthy open source community: people, dialogue, and turnoverArrow as critical infrastructure: DuckDB, Polars, Pandas, and the spec that unifies themManaging 4,000 open issues without losing your mindImposter syndrome in open sourceWhat the yoga mat teaches you about discipline and collaborationAI and the future of programming: 100x more software or 10x better software?

Emily Riederer writes Python with an R accent, and we’re all comfortable with it. In this episode, Emily reflects on her journey through R, Python, and SQL — from lessons learned in averaging default values (oops, we're not all rich!) to discovering that column selectors are way cooler than they sound. She weighs in on the delicate art of learning in public, why frustration often makes the best teacher, and how to find your niche by solving the boring problems. Oh, Oh, and the crew casually drops that she's keynoting posit::conf 2026!Episode NotesEmily’s had a wild ride through modeling, data engineering, machine learning, and back again, and she knows a thing or three about the evolution of SQL tooling (from nightmare multi-page scripts to the dbt renaissance). She reveals how building internal packages became her gateway to making work enjoyable. Plus: the surprising Stata origins of column selectors, the eternal struggle of naming packages across R and Python, and why watching people code teaches you more than any tutorial ever could. The conversation gets real about imposter syndrome and the magic of tacit knowledge.What’s InsideWhy real-world data is chaos, not truthThe path from modeling to data engineering (and back)What a data pipeline really is (extract, load, transform) and why organization mattersHow dbt changed the SQL game Learning by watching: Tacit knowledge and coding over the shoulder Imposter syndrome and learning in public Building internal tools to escape busyworkposit::conf 2025 keynote preview

Rebecca Barter, senior data scientist at Arine and adjunct assistant professor at the University of Utah, refuses to work on things she doesn’t care about. Lucky for us, she cares about a lot, most of all impact. In this episode, Rebecca joins The Test Set to talk about learning fast, building better tools, and staying motivated and adaptable.She shares how moving between R, Python, SQL, and dashboards reshaped how she thinks about expertise. Plus a reflection on her recent posit::conf talk, “AI: Hype, Help, or Hindrance.”Episode NotesRebecca digs into what it’s really like to work with AI every day and why humans still rule, especially in exploratory data analysis. She explains how tool building can be the fastest way out of busywork and how teaching beginners sharpened her ability to communicate clearly. The conversation circles a bigger question too: As AI keeps improving, are we headed toward a future where code looks completely different … or maybe disappears altogether?What’s InsideWhy motivation matters even more than productivityEscaping busywork by building better toolsFrom R to Python to dashboards: Learning fast as a survival skillReality check on AI in the IDEWhy exploratory analysis still needs human intuitionThe 80/20 of coding: Automate the boring, protect the judgmentTeaching beginners and earning trustThe uncertain future of code

You’ve probably used Narwhals without realizing it. It’s the compatibility layer helping apps and libraries like Plotly play nice with Pandas, Polars, Arrow, and more — while keeping computation native instead of converting everything to Pandas. In this episode, Marco Gorelli explains how his weekend experiment turned into essential ecosystem infrastructure and why data types, not APIs, are where interoperability gets tricky. Plus what it takes to build trust and community around an open-source project. Episode NotesMarco shares the Narwhals origin story (including the meme-powered name), the hard edge cases that live in data types and null semantics, and why he’s cautious about using AI for code generation when correctness hinges on tiny details. We also jam on proactive “GitHub surfing,” conference talks as trust-building exercises, celebrating contributors, and how early commit messages capture the genuine excitement of building something new.What’s InsideNarwhals 101: You’ve probably used it (even if you didn’t know it)The real interoperability traps: data types, null semantics, and “looks-the-same” operationsWhy expression systems won, and how they shaped Marco’s approach — with nods to Ibis, Polars, and PandasOpen source as social work: proactive outreach, trust-building, and a Discord-powered communityExtending Narwhals to new engines, starting with the Daft plugin

In this episode, we’re joined by Kelly Bodwin — candy corn defender, board game enthusiast, and Associate Professor of Statistics and Data Science at Cal Poly. We discuss her path from English and French to statistics, how she builds teaching tools and navigates AI in the classroom, and what it takes to keep a programming community weird in the best possible way.Episode notesKelly is curious, collaborative, and unafraid to lean in on quirky. Kelly shares how she balances teaching three courses with master's student supervision, applied research projects spanning Polish history and beyond, and her belief that the best part of academia is the people. We also dive into the practical and philosophical challenges of staying current in a field that reinvents itself every few years.What's insideBreakfast mixologyBuilding Quarto extensions with JavaScript and AIWhen ChatGPT helps students learn (and when it doesn't)Applied stats meets history: analyzing social networks from the Polish RevolutionWhy remarkable, welcoming communities matter more than perfect code

This episode is Part 2 of our conversation with James Blair. He explains how he found his “accidental perfect fit” as a solutions engineer and how that role became a pipeline into product management. Get a peek into the AI-powered tooling he’s now building for the Posit ecosystem, and hear how he’s using Claude Code, Positron Assistant, and DataBot to generate synthetic, industry-specific demos on the fly — plus, why the real magic is keeping humans firmly in the loop. Episode notesThis is a story about listening deeply to users and using AI to make that listening scale. James explains what solutions engineers actually do, how that work shaped Posit’s product team, and how synthetic data plus agents are changing the way they build demos and teach data science. What’s insideWhat a solutions engineer really is and why the role was such a good fit for JamesHow solutions engineering became a natural pathway into product management at PositMulti-agent “bot posse” workflows and why context management mattersUsing AI the right way and why code literacy, critical thinking, and staying human are the real superpowers in an AI-saturated world

In Part 1 of our conversation with James Blair, we trace his delightfully non-linear path from childhood robotics dreams to journalism to R, with a few stops in between. We hear about the Shiny app that changed his career, plus a candid roundtable with Michael, Hadley, and Wes about whether a data-science master’s still pays off in the age of AI.Episode notesThis is a story about staying hands-on and fiercely inquisitive — whether analyzing bike telemetry or in teaching data science. James shares how early experimentation with Shiny helped shape his career, and how curiosity (not credentials) still powers meaningful work in data science.What’s insideA winding path from robotics to journalism to psychology to data scienceDiscovering the power of applied statsThe value (and limits) of a data-science master’s in a shifting AI landscapeFighting confirmation bias: good analysis resists the answer you want

In part two of our conversation with Julia Silge, we discuss how work actually ships: the boundaries, the glue, and the tools that turn noise into signal. From there, we go macro and wonder what the LLM era means for humanity’s contributions, plus how licensing is evolving to protect sustainability without abandoning openness.Episode notesBoth practical and philosophical, this conversation spans workplace energy, team connective tissue, and the big questions LLMs have us asking in a shifting data science landscape.What’s insideJulia’s system for turning scattered community signals (GitHub, Stack Overflow, discourse) into product insightThe power of “glue” work, and where to find the winsFrom Stack Overflow to LLMs: What changed when communal Q&A became model fuel — and what that means for finding answersLicenses in a new era: Threading the needle between MIT-style generosity and elastic-style sustainability for platformed softwareTry Positron: Where to download, read docs, and give feedback

In part one of our conversation with Julia Silge, astronomer-turned–data-science leader, we explore why data science needs a different kind of IDE. Julia takes us inside Positron, Posit’s next-generation, data-scientist-first environment, and unpacks the day-to-day realities that make data science work unlike software engineering. Along the way, we get a first-hand account of a legendary pineapple-pizza protest and how to juggle multiple projects at once.Episode Notes:A behind-the-scenes tour of Positron and the workflows it’s built for, plus the stories, trade-offs, and team choreography required to ship an IDE on a living substrate. We talk extension ecosystems, upstream merges, data viewers, and more. Plus, Julia shares why applied systems (and messy, real-world data) are her happy place.What’s Inside:The pineapple-pizza story that unexpectedly went viral — and what “context collapse” feels like from the insideWhy Positron is a data-science-first IDE, optimized for analysis, not general software engineeringIteration vs. reproducibility: the central tension in data science workflows and how tooling can honor bothHadley’s cold-turkey move from RStudio, muscle memory, and finding the new ergonomic grooveHow Julia measures success by smoothing the boundaries between tools and teamsThe applied, people-and-process side of data science that keeps Julia energized

For this episode, we turn the mic around. Wes McKinney takes over the interviewer’s chair to chat with his co-host, Michael Chow. Michael’s a principal software engineer at Posit, but he started out studying how people think — literally, with a PhD in cognitive psychology. Somewhere along the way, he got hooked on data science, helped build adaptive learning tools at DataCamp, and now spends his days thinking about how to make Python easier to use and more fun.The two dig into what drives Michael’s curiosity, how a “weird obsession with tables” turned into a beloved open source project, and the future of data science/scientists.Episode Notes:We explore Michael’s path from studying the mind to shaping the Python data science ecosystem. From adaptive learning platforms to Great Tables, Michael shares how following unexpected curiosities can spark tools and communities that last.What’s Inside:Michael’s pivot from an academic career in data scienceBehind-the-scenes messiness of building data and learning platformsOpen source projects born out of zany, single-minded passionsBringing beauty to rows and columnsBig-picture thoughts on where data science — and open source tooling — are headed

Michael, Hadley, and Wes welcome Roger Peng, professor of statistics and data science at UT Austin and co-host of Not So Standard Deviations. Together they trace Roger’s journey from early R adopter to pioneering online educator and prolific podcaster. The conversation ranges from the accidental rise of “data science” as a field, to the tension between research papers and software maintenance, to what makes for meaningful, lasting creative work.What’s Inside:Roger’s first analysis project and what it taught him about authorship and dataRoger’s advice for students testing the waters in data scienceWhy software has become the unifying language of modern statisticsThe origins of “data science” as a field and a labelReflections on Coursera, MOOCs, and opening education to the worldWhat keeps a podcast (and a career) going strong after a decade-plus

In this conversation, Mine Çetinkaya-Rundel, data science educator at Duke University and Posit, joins Michael, Hadley, and Wes to talk about teaching data science in a time when AI can write the code for you. Mine shares her journey from actuarial science to academia, the teaching philosophy behind the “whole game” approach, and her experiments using LLMs for instant student feedback. Along the way, the group dives into the joys and risks of coding by hand, the role of open source in the classroom, and what it’s like to work across both the R and Python communities.What’s Inside:How a career in actuarial science led Mine to the world of data science and teachingThe “whole game” approach to learning and how it helps students stay motivatedBuilding an LLM-powered feedback tool for low-stakes assignmentsBalancing AI assistance with the need for hands-on coding experienceThe shared DNA of R and Python scientific computing communitiesThe hidden value of live coding, pair programming, and seeing the process — not just the output

In part two of our conversation with Wes McKinney, we dig into the challenges and realities of sustaining open source development. Wes shares how funding actually works (or doesn’t), why corporate buy-in is essential, and what it’s like building tools across languages, communities, and IDEs. We also talk about the Apache Software Foundation’s role in open governance and the origin of the Positron IDE.What’s Inside:Why passion isn’t enough for open source to scaleApache Arrow’s origin story and how it was pitchedHow open governance enables trust between competitorsThe thinking behind Positron, Posit’s next-gen IDEPolyglot programming – Designing tools that bridge the R/Python divideLLMs and data UX: Why modern IDEs need to serve both humans and modelsDay-to-day coding, advising, investing, and context-switchingMetalheads unite

Wes McKinney’s fingerprints are all over the modern data stack — from inventing Pandas to co-creating Arrow. But before all that, Wes was organizing speedrun communities and hacking together better ways to wrangle datasets in finance. In this conversation, he shares his origin story and what makes good tools good. Stay tuned for part 2, coming soon.What’s Inside:How frustration with data work led Wes to build pandas (and leave a PhD)A nostalgic dive into the GoldenEye speedrunning sceneWhy read_csv performance is a deeply personal crusadeLessons from convincing friends to quit finance and go open sourceFounding startups, launching Arrow, and the Ibis origin storyThe beauty of letting contributors take the reinsShout-out to Philip Cloud, pandas’ resident pun masterWhy open communities win — and what it takes to build them

Before Hadley Wickham became a pillar of modern data science, he was a spreadsheet-loving teenager making databases for his dad’s job. In this episode, he reflects on the early days of his involvement with R, the birth of tidyverse, and how real-world unpredictability — like a bear in a field — shapes data science.What’s Inside:Hadley’s first brush with R code … inside a Word docConsulting as a grad student — and learning what people really want from statsHow messy Excel sheets inspired the tidy data revolutionWriting R packages as a form of self-defense (and productivity)The secret sauce of building the tidyverse teamOn focus, burnout, and saying “no” to GitHub pull requestsCurrent obsession: using LLMs  to make data science faster, easier, and more funHow writing books is a form of tidying ideas, and how a Shiny textbook led to a custom bike