Jan van der Vegt

Introducing Exa Monitors - your agent’s radar for the web Exa is a search engine built from scratch, and today we're exposing our "update" layer. Simply define what to find and how often - Monitors will return any new information, via webhook.

Is a non-alcoholic someone who is not addicted to alcohol or someone who is addicted to non-alcoholic drinks?

Introducing Exa Deep: putting an agent inside every search For each query, an agent runs in a loop until it gathers all information, then returns structured output. Evals show Deep is Pareto optimal at 4-60s latency, ideal for quick, cost-efficient research!

We made Valentine's Vectors, an AI app that finds people similar to you from the web :) Uses Exa search, try it here - exa.ai/valentines

@officialKrishD @ExaAILabs Hit me!

@ExaAILabs You guys open for - “open feedback” ?

Introducing Exa Instant: the first sub-200ms search engine. Faster than Google, it's custom built to power realtime AI products like chat and voice.

See your X Wrapped 2025! Exa-powered AI analyzes you, roasts you, and even predicts your future 🔥 Get yours here: exa.ai/xwrapped

Here's my X Wrapped 2025: exa.ai/xwrapped/dasdzy x.com/ExaAILabs/stat…

I want to have a secure personal AI agent that can connect to my email, calendar and possibly some e-commerce accounts or WhatsApp. What are my best options? I want it to be as good as possible, don't mind spending for top models

As bad as the state of LLM evals is, the state of search evals is even worse. That's why along with today's improvements in People Search, we're also releasing an opensource eval dataset to become a benchmark others can use. Our hope is that these types of search evals will jumpstart more progress in retrieval research. We can all only climb hills if there exist hills to climb. Luckily, after decades of stagnation from traditional search algorithms, there are now many untapped hills. People search is just one example, one that's particularly helpful for recruiting, sales, and even just making this big world feel a bit smaller. More evals to come :)

@DavidSHolz I frequently organize cube drafts with tech people in SF, you are welcome to join the pool! cc @jeffr

any friends or moots play magic the gathering? and live in the bay area?

@OndrejStrasky San Francisco is not like the rest of California :p how long are you in town for?

Another trip over the ocean. This time San Francisco for GDC. My first time at the conference! I was super excited to visit sunny California, but I just looked up the weather and now I feel scammed 😭

@markovictory Unfortunately I'm settling a bet and losing

@DasDzy My head canon is that you're writing the vows for a couple of card gamers at a shotgun wedding.

Quick question, when shuffling two piles of cards into each other - do you pronounce that as "riffle" shuffling or "rifle" shuffling?

Why is MODO so laggy?

R is feeling shy today

Honored to join ⁦@DasDzy⁩ for the inaugural draft of his cube. *And* he made dinner. What a guy! ❤️

@ecobaronen Ponder

What's the pick to start off this high stakes draft?

540 card artifact cube with just 4 colors. Good amount of colorless cards and artifact-based fixing, but also decent amount of multicolor cards. How many fixing lands would be optimal? Doing the first draft tomorrow!