Dinesh K Visvanathan

🎓CS | 👨‍💻 Building @getdini
🧬 Solving health problems with AI
🌱 Built @dinihabits during burnout

[Dinesh K Visvanathan @dini_kv]

Last year I stood on this stage pitching @getdini after graduating from the @Communitech startup bootcamp.

#throwback

[Major League Soccer @MLS]

WHEN ARGENTINA NEEDED HIM MOST.

LIONEL. ANDRÉS. MESSI. 🐐

[Dinesh K Visvanathan @dini_kv]

Happy Canada Day 🇨🇦🍁

[Dinesh K Visvanathan @dini_kv]

Composer 2.5 is far better than previous composer models. Good improvements by team @cursor_ai

[Dinesh K Visvanathan @dini_kv]

History is about to be rewritten ❤️😊

[Dinesh K Visvanathan @dini_kv]

Most people use AI, very few understand it. This will change that. #AI #BTS

[Dinesh K Visvanathan @dini_kv]

Your AI just confidently lied to you. Here's why, and how to stop falling for it.

[Dinesh K Visvanathan @dini_kv]

I wish someone gave me this when I started. So I made this for founders who don’t want to learn the hard way.

[Garry Tan @garrytan]

There is no proper eval for personal brains on OpenClaw and Hermes Agent

If someone wants to build this, this would be awesome. I'll build my own version inside of GBrain but I think it would still be useful to have a 3rd party build this to properlye valuate this new category

[Dinesh K Visvanathan @dini_kv]

I just 10x'd my coding flow. Built a local-first Wisprflow alternative for my own workflow (honestly better than Wisprflow) and it's ridiculously fast. No API keys. No latency. No cloud processing. Just works.

Should I open source this? 👀

[Dinesh K Visvanathan @dini_kv]

Gemma 4 is best local replacement for your chat model, its impressively smart. Agentic and tool calls are not up-to the mark.

[Dinesh K Visvanathan @dini_kv]

I have been using @googlegemma Gemma 4 locally since it launched. Wow! I am impressed with the results.

[Dinesh K Visvanathan @dini_kv]

💛💛💛

[Dinesh K Visvanathan @dini_kv]

I have been using @googlegemma Gemma 4 locally since it launched. Wow! I am impressed with the results.

[Dinesh K Visvanathan @dini_kv]

Local AI is winning 🏆

[Dinesh K Visvanathan @dini_kv]

Local AI is winning 🏆

[Dinesh K Visvanathan @dini_kv]

Apple said better luck next time 🥲

#WWDC26

[Dinesh K Visvanathan @dini_kv]

Please, @Apple let me in this time! #WWDC26 🥺

[Dinesh K Visvanathan @dini_kv]

@aroussi 😂😂

[Dinesh K Visvanathan @dini_kv]

@TheGeorgePu How about today😀

[Google AI Developers @googleaidevs]

https://t.co/pfrSbEEYrH

[Dinesh K Visvanathan @dini_kv]

Google just shipped Silicon Valley’s middle-out compression IRL 😂TurboQuant = 6x smaller KV cache, 8x faster, 0 accuracy loss. The GIF is the Pied Piper demo, frame for frame.

Weissman score just went nuclear. Pied Piper won.

[Google Research @GoogleResearch]

Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: https://t.co/CDSQ8HpZoc

[Dinesh K Visvanathan @dini_kv]

@ftnabeelah @benln @shayaan_azeem @demireren_ @byjustinwu thnx :)