Hex

Data teams are becoming context curators. As AI adoption grows, it’s clear that teams need to have an eye on the context that helps direct agents to the right answer. Meet Context Suggestions: Hex monitors your agent interactions, spots where better context would improve answers, and automates the fix. 🧠Your data agent gets smarter every time someone uses it. Learn how below ⬇️

✨ANOTHER PM HAS ENTERED THE VILLA........MEET ANDREW✨ What you need to know about Andrew: >He enjoys a good pair of pants >His favorite food is egg >His current hobby is skateboarding >He cannot do a kickflip, please don't ask He also thinks Hex should get smarter every time you use it. Thankfully, it does. More about Context Suggestions here ➡️ bit.ly/4cTKvEb

PMs aren't writing PRDs anymore. @braintrust 's @ameyab on why the new PM superpower in the age of AI is evals. TL;DR: Find where your AI fails → build better evals → ship with confidence.

@barrald at @_hex_tech on their mission towards making everyone a data person, from sales reps to executive chefs.

📍Get started with Context Suggestions: hex.tech/blog/introduci…

✨CHARLENE HAS ENTERED THE CHAT✨ Some critical Charlene lore: >She's a 2x Salida Union School District spelling bee champion >Her fave snack is dehydrated chicken chips (proteinmaxxing ftw) >If she had a mascot, it would be a slinky >She thinks data teams need to stop second-guessing the agent's work and start using Review Agent Learn more about Hex's Context Suggestions and how you can get started below ⬇️

📍Your data agent is now smarter: hex.tech/blog/introduci…

Data teams are becoming context curators. As AI adoption grows, it’s clear that teams need to have an eye on the context that helps direct agents to the right answer. Meet Context Suggestions: Hex monitors your agent interactions, spots where better context would improve answers, and automates the fix. 🧠Your data agent gets smarter every time someone uses it. Learn how below ⬇️

We're teaming up with @OpenAI for a panel and happy hour on Product Engineering in the Age of AI. Hear from @amoljain_ of @Replit; @DeeniFatiha of @GammaApp; @rohanvarma of @OpenAI and our very own @dseisenberg on the future of product engineering and what building great products looks like in 2026. 📅 April 29 | 5:30–8PM | Hex HQ Spots are limited — request to join: luma.com/yvleukg7

The @_hex_tech agent has gotten really good. I give it a list of tables and a detailed breakdown of the analysis I want. It generates a SQL/python notebook with charts within minutes Great for exploratory work. I ask it to slice the data one way, tweak my assumptions, tell it to adjust the logic, check the results, and iterate my way to the insight I need. Same workflow I've been doing for the past six years, just faster because I'm chatting instead of writing all the code myself Claude Code + Codex can do this too, but what makes Hex stand out for this use case is that all your SQL queries are saved by default for future reference, and you can instantly publish the notebook charts as a live dashboard

The power of Hex, directly in your terminal: hex.tech/blog/introduci…

✨LIVE FROM HEX'S MIDWEST SATELLITE CAMPUS (Hogan's House) — IT'S HOGAN ✨ What you need to know about Hogan: >His high school superlative was “most inspiring” >He thinks all dogs should have human names >He loves the terminal so much his dotfiles have over 12k lines of code >All of you are invited to his wedding if you try the Hex CLI (bit.ly/4mzgKgz) (💥And to learn more about the Hex CLI—check out the link below ⬇️)

Join us: hex.tech/careers/

"It's the tech enthusiast that's winning." @baseten's @DannieHerz on what tech leaders should actually be looking for when building teams. (P.S. We're hiring tech enthusiasts at Hex, link below ⬇️)

I wrote a letter to the next PM @_hex_tech about what it's like to work at Hex right now. If i had to distill it to one line, this is it. If this resonates, I'd love to chat.

Hex is now a native MCP app in @glean 🤝 Glean knows your company. We know your data. Together, we're able to give your AI the full picture — blending organizational context from Glean's Knowledge Graph with our trusted, curated data insights to power a richer analytical experience. Whether it's our CLI, API, Slack, Native Apps, or inside Hex we're excited to keep bringing our agentic experience closer to your existing workflows.

@hamiltonulmer chopin's "nocturne in e minor" and then an encore of "chopsticks"

@_hex_tech what's Claire playing at the recital

✨MEET CLAIRE✨ Some fun facts about Claire: >She has a part-time dog >Her second ever piano recital is next month (comment below for the invite) >She took 353 Lyft bike rides last year >She thinks the Hex CLI is the best way for data teams to work in 2026 Learn more about Hex CLI and how you can get started: bit.ly/4t9C1Qw

🎙️Introducing Max Agency Max Agency is a new podcast where we go deep on how the best agents are actually being built: architecture decisions, tradeoffs, evals, and everything in between. Each episode, I sit down with engineering leaders who are doing this work in production. Our first episode features Izzy Miller (@isidoremiller), AI Engineer at Hex (@_hex_tech). Hex has been shipping data agents since before most teams were even thinking about them, starting with single-cell text-to-SQL and graduating to a full Notebook agent that can work autonomously for 20 minutes on a complex analysis. Izzy has a lot of perspective on what it actually takes to get agents working well in production, and what breaks along the way. A few takeaways from our conversation: - Keep your eval sets small enough to hold in your head: Izzy runs 30-50 handcrafted "traps" with multiple repetitions, rather than hundreds of variants. If you can't explain why your agent fails each one, your eval set is too big - Day zero performance is almost irrelevant: The more interesting question is how the agent compounds. Izzy is building a 90-day simulation where the warehouse evolves and the agent has to accumulate understanding - You can catch agent errors without seeing the raw outputs: By running an LLM-as-a-judge over production usage and clustering the results, you can surface places where something likely went wrong, without needing to read individual conversations Watch the full episode on: - Youtube: youtube.com/watch?v=Xyh1Eq… - Apple Podcasts: podcasts.apple.com/us/podcast/how… - Spotify: open.spotify.com/episode/1BJlg3…