Utkarsh Varman

ASR @SarvamAI | Ex-GenAI @SamsungResearch | Ex-@ShiprocketIndia | Deep Learning 🪢

[Jay Yang @Jayyanginspires]

CEO of Stripe's advice to young ambitious people:

[kivi @heykiviai]

kivi for Mac is here.

Typing is officially optional on your Mac.

Talk in any app, and kivi turns what you say into clean, ready-to-send text. No more typing out long emails, messages or prompts.

Here's everything kivi can do 🧵

[Utkarsh Varman @UtkVar03]

Dopamine rush for the weekend

[Tibo @thsottiaux]

Hi Astra users. A reset and a quick update on quality issues that have been posted around.

Working with some of you, we have found and fixed the following issues:
- Some skills written for previous models were triggering too often or preventing the model from checking its work.
- An opt-in context management experiment that could cause early stops or replies to older messages. We've disabled it. Our rough estimate is that 4-5k users were affected by this experiment.
- We've also removed some badly configured engines that resulted in a measured quality degradation for a long tail of traffic flowing through them.

We’ve also made some more minor improvements and things should feel significantly better across the board. More consistent follow-through, better tracking of your latest message, and better checks on the work as it’s going through the motions.

The examples posted and all the users who worked directly with us were incredibly useful in helping fix things quickly. Always grateful for this incredible community.

And of course, a reset is also landing by midnight today.

[Pratyush Kumar @pratykumar]

📢 Open-sourcing the Sarvam 30B and 105B models! Trained from scratch with all data, model research and inference optimisation done in-house, these models punch above their weight in most global benchmarks plus excel in Indian languages.

Get the weights at Hugging Face and AIKosh. Thanks to the good folks at SGLang for day 0 support, vLLM support coming soon. Links, benchmark scores, examples, and more in our blog - https://t.co/DcCG3zlN8p

[Anshu Sharma 🌶 @anshublog]

Totally in awe of @SarvamAI team.

When I met @vivekrag at KV Summit, I was highly skeptical of the space and what could be done here uniquely - even as I admired Vivek and @pratykumar for their prior work.

They have totally killed it so far.

Onward to a winning team.

[Ashwini Vaishnaw @AshwiniVaishnaw]

Connected with the enthusiastic youngsters, full of ideas and optimism about India’s AI future.

📍AI Impact Summit, Bharat Mandapam

[Sarvam @SarvamAI]

Join us for the next drop. This time, in-person.

Date: February 18, 2026
Time: 12:30 PM
Venue: Bharat Mandapam, Plenary Hall B

[Utkarsh Varman @UtkVar03]

Yep! We got our own smart glass😎

[Pratyush Kumar @pratykumar]

Drop 12/14: Models, products, impact - today something different, very different. Launching Sarvam Kaze, our foray into getting our models into the your hands with our devices - designed and built here in India!

[Pratyush Kumar @pratykumar]

Drop 12/14: Models, products, impact - today something different, very different. Launching Sarvam Kaze, our foray into getting our models into the your hands with our devices - designed and built here in India!

[Aditya Dhavala @dvs_aditya]

India builds with @SarvamAI

[Sarvam for Developers @SarvamForDevs]

All in for the big launch 🚀

[Sarvam for Developers @SarvamForDevs]

All set for something big at IMPACT India AI Summit.

Hall 2.
Ground floor.

Drop by our booth.

Get the sneak peek
before the official drop.

[Utkarsh Varman @UtkVar03]

@selfawareatom Letss goo!!!

[Utkarsh Varman @UtkVar03]

After months of hard work and sleepless nights, it’s finally out! Proud to introduce Saaras v3 our latest Indic ASR model supporting 22 indic languages.
Check it out - https://t.co/9Ca56xC724

[Pratyush Kumar @pratykumar]

Drop 8/14: Introducing Saaras V3, the next iteration of our speech recognition model. We extend our lead in this space with an even more accurate model, particularly for mixed-language and noisy speech.
We have also expanded support for all the 22 scheduled languages of India.
And we now support real-time streaming! In this mode, the model delivers low latency while preserving transcription quality.
https://t.co/SQSgCPsUr5

[Pratyush Kumar @pratykumar]

Drop 8/14: Introducing Saaras V3, the next iteration of our speech recognition model. We extend our lead in this space with an even more accurate model, particularly for mixed-language and noisy speech.
We have also expanded support for all the 22 scheduled languages of India.
And we now support real-time streaming! In this mode, the model delivers low latency while preserving transcription quality.
https://t.co/SQSgCPsUr5

[Pratyush Kumar @pratykumar]

Drop 5/14: Introducing Bulbul V3, our latest text-to-speech model. It raises the bar for how human it sounds, while being super robust.

In an independent third-party human listening study, Bulbul V3 delivers the highest listener preference, and low error rates across use-cases and languages. See details in our blog, but first watch the video. https://t.co/HkijWrupi9

[Sarvam @SarvamAI]

The bird you know. The voices you don't. YET!

[Utkarsh Varman @UtkVar03]

@hustle_fred Thank you Fred! 🙌🏻

[Utkarsh Varman @UtkVar03]

Kicking off Day 2 at #100xBuildathon2!

We’re building Vocaliq AI — a platform designed to bring collaborative music generation with AI.

Wrapped up basic frontend today. Next up: backend integration. Stay tuned as we keep building!

#buildinpublic #GenAI #100xEngineers

[Utkarsh Varman @UtkVar03]

What is your definition of being skilled in 2025?

Being able to write majority of code from scratch?
Or
Being able to understand and refine code written by AI?

#AI #Automation #AgenticAI #JobHunt