alexis

in love with the act of creation

i wrote a substack about this months ago: "Because error compounds, small reductions in [agent] error rates have out-sized benefits. And that’s exactly what happened." read "Accumulated Drift" - link below

@kesava_kirupa @snarkyzk @luminai huge!!!

America’s leading health systems, like the Cleveland Clinic, work with @Luminai to eliminate administrative waste. We’re rapidly deploying to more health systems, and excited to announce Series B, bringing total funding to $60m.

@_Dave__White_ @paradigm @matthuang @alanapalmedo @danrobinson @FEhrsam incredible run :)

today was my last day @paradigm i'm so grateful to @matthuang, @alanapalmedo, @danrobinson, @FEhrsam, and the rest of the team for the opportunity and for taking such good care of me over the past five (!) years i'm also very excited for my next chapter. watch this space.

@rsdgpt 😂

Claude Mythos is just Claude Opus on gstack.

@saraareynolds 🔥

@aditabrm huge!!

Today, we’re setting the new standard for complex document extraction. Introducing Deep Extract. It utilizes an agent harness approach that repeatedly iterates and verifies outputs until they are at human-level accuracy. More below:

NASA has released the first image of the entire Earth in nearly 55 years. The photo was taken by astronauts aboard the Orion spacecraft. The northern lights are visible at the pole. The last image of Earth from deep space was taken in 1972 during the Apollo 17 mission; all images from the past 50 years are a composite of multiple images from low orbit.

@saraareynolds lmao iconic

this weekend I was at silly hacks and I vibe coded something everyone definitely needed: a restaurant rating app this one just happens to rank places by their bathroom turns out the bathroom tells you everything u need to know check it out porcelainpalace.club built with @sendbluehq for notifications @Replit for a lil animation

@maxleiter @shuding @shadcn @jaredpalmer @gaspargarcia_ @v0 what a run!!

After almost four years, I’ve made the difficult choice to leave Vercel. It’s been quite the ride; from starting the AI SDK and v0 with @shuding, @shadcn, and @jaredpalmer to helping build and lead the v0 team with @gaspargarcia_. I’ll miss everyone, but especially the @v0 team — I wouldn’t want to build it with anyone else. I’m taking a few weeks off, then joining something exciting. Looking forward to sharing more soon.

@saraareynolds You might like what @selinayfilizp is thinking about! x.com/selinayfilizp/…

so nerdsniped by this rn. my random idea that i want to explore tangentially is if simulations are more powerful if they are operated on "real agents" by "real agents" i mean agents that are actually impressioned by their humans bottleneck here is rules around personal agent data sharing, memory, & context.

@selinayfilizp love the stress test idea

AI agents have SOUL.md for personality. SKILL.md for capabilities. AGENTS.md for rules but when your agent faces a tradeoff (act vs. ask, spend vs. save) it has zero framework for choosing so last weekend, I built an interface that helps you train your agent on how you decide

@lennysan @raindrop_ai in San Francisco! Just DM me to apply :)

Who’s hiring engineers right now? Reply with the role, location, and how to apply.

being a founder is weird because everything will be normal and then someone will randomly dress up like alexander hamilton and start rapping about your company

recurring revenue graph for @raindrop_ai is approaching a vertical line

The best way to celebrate 1 year of Responses API is to take a look at what you've shipped. @raindrop_ai is using it to monitor agent behavior in production.

average day at @raindrop_ai

@heymikasagi @raindrop_ai @langfuse @LangChain @RespanAI @PortkeyAI @braintrust @helicone_ai @mintlify @arizeai very cool!

New evals on Agent Arena: ✨LLM observability✨ AX (agent experience) ranking as of 3/03/36: 1/ @raindrop_ai 2/ @langfuse 3/ @langchain 4/ @RespanAI 5-6/ @PortkeyAI, @braintrust 7/ @helicone_ai (now part of @mintlify) 8/ @arizeai More on 2027.dev/arena We measure how easy it is for AI agents to get started with devtools, fully autonomously With AI agents becoming the primary consumers of docs and APIs, AX is the natural evolution of DX DM me for your full AX eval! If you're missing a tool or category, comment below cc @benhylak, @alexisgauba, @snarkyzk, @marcklingen, @maxdeichmann, @nimarblu, @hwchase17, @ankush_gola11, @samecrowder, @Andydy42, @raymond_huang26, @jumbld, @ankrgyl, @daRubberDuckiee, @justinstorre, @coleywoleyyy, @jason_lopatecki, @seldo