Sandhya

Co-founder @Calibre_Labs | Applied AI research & consulting | Agents, AI Evals | prev EVP @amplitude_HQ | VC @khoslaventures @sequoia | @stanford @iitbombay

[Sandhya @sandhya]

https://t.co/uAxtFPV5mC

[Sandhya @sandhya]

@HamelHusain @sh_reya Chief Looker at Data

[Sandhya @sandhya]

@swyx @jediahkatz @poteto why do we have to declare everything dead to discuss their evolution 😞

[Sandhya @sandhya]

AI long form writing with a good skill/examples has now gotten good enough with a few models that i actually like having it to do the final-but-one draft.

in a few months (weeks?) pangram scores will be irrelevant. we will still see slop but wether AI was used will no longer be the right flag.

[Sandhya @sandhya]

@evancharles this post requires more explanation evan

[Sandhya @sandhya]

currently free on ollama, opencode and nous

[Sandhya @sandhya]

love seeing a new open source success 🫡🇺🇸

[Reflection @reflection_ai]

We just served 1 trillion tokens of Beam in 24 hours. To celebrate, we've reset everyone's limits.

Go build something big 👉 https://t.co/ry8O6JPlz6

[Sandhya @sandhya]

congrats Hamel! you’ve been such a big inspiration for me and @justinjbauer as we work with clients on AI ❤️

[Hamel Husain @HamelHusain]

👀

[Sandhya @sandhya]

@spenserskates wow congrats to both of you!!!

[Sandhya @sandhya]

$META is going to have a good q4

[Alexandr Wang @alexandr_wang]

introducing muse wonder

our new frontier model. state-of-the-art at personal agents and professional agents.

our largest and most capable model yet—rivals the big bois

[Sandhya @sandhya]

@nicbstme 💯

[Sandhya @sandhya]

returned a dress
frill issue

[lauren @poteto]

fish dead
gill issue

[Sandhya @sandhya]

product release videos got fully automated at the same time grokbot started monitoring twitter for us.

nobody can say this timeline doesn’t have a deep sense of irony.

[Sandhya @sandhya]

🙌 no eval rubric == no product strategy in AI

[Jerry Liu @jerryjliu0]

These days, you can pretty much solve any task by defining an eval and hillclimbing over it, instead of directly defining the deterministic/agentic workflow to solve it.

The data provider companies' entire job is to define evals for all economic activity to make the frontier models capable of doing anything.

Your job then becomes pointing frontier intelligence in the right direction - defining what to solve, and how to measure what good looks like.

Agent application interfaces will evolve to capture this. The most complex processes will still need some sort of explicit workflow builder interface, but most tasks can be compressed into goals and eval instructions.

[Sandhya @sandhya]

https://t.co/ZAL9KeO3Lb

[Jason Zhou @jasonzhou1993]

Introducing OpenSource Clay killer

- 85% cheaper than clay
- Most accurate on people-search bench
- 25x faster
- Run GTM directly in claude code or codex

Try it at https://t.co/s60Ye4EO3h

Repo link below 👇

[Sandhya @sandhya]

Big day for 🇺🇸 OSS 🫡

[Reflection @reflection_ai]

Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active.

- Frontier reasoning efficiency
- Advances the Western open frontier on coding & agentic tasks
- Trained end-to-end from scratch

Full weights release this month.

Learn more about Beam: https://t.co/c3Qx2cpM8G

[Sandhya @sandhya]

@ajambrosino @reach_vb thank you

[Sandhya @sandhya]

one side effect of agentic work getting very good is it allows me to think a lot more about what i want to build without worrying about the deadline coming up. the execution of well articulated ideas (with context and artifacts ready to go) is 100x faster.

i can’t overstate how much i am enjoying this as a professional procrastinator when i am still searching for perfect clarity 😁

[Sandhya @sandhya]

@Ascion_Next @tszzl maybe they are optimizing for people standing close to it/under it. michelangelo also got the same grief from critics ;)