Nikhil Shah

Building @slidelyAI (YC S22)

[Nikhil Shah @iamshnik]

@yousifa While building the sandboxed agents at slidely, we used our own hosted proxy to send requests to llm providers, and authenticated it both sandbox id, and short lived tokens on the sandbox that we would rotate on every message

[Nikhil Shah @iamshnik]

@tetumemo Definitely give https://t.co/fwpKTfgMSa a try if you like that!

[Nikhil Shah @iamshnik]

@_simonsmith Give https://t.co/fwpKTfhkHI a try, it can take inspiration from nano banana and create fully editable powerpoint presentations

[Jaxmatrix @Jaxmatrix2]

@GithubProjects Well if you are aiming for genAi slides probably should use https://t.co/wQdfBbsbt9 its output is beautiful that just works with powerpoint (and yes actual powerpoint shapes that can be tweaked) also powered by Nano banana

[Nikhil Shah @iamshnik]

@airtelindia @airtelnews How do you charge for the same service TWICE IN THE SAME TIME PERIOD, ESPECIALLY WHEN I WAS CONFIRMED THAT IT IS CANCELLED AND I WONT BE CHARGED FOR IT

@airtelindia @airtelnews

[Nikhil Shah @iamshnik]

@airtelindia @airtelnews has the shittiest support ever, I swear.

No way to reach the customer care.

No way to cancel Xstream.

Last month I somehow managed to get in touch with the customer care and they told me that they have cancelled the Xstream plan, instead...

[Nikhil Shah @iamshnik]

Mai kabhi batlata nahi, par static current se darta hu mai ma 🥹

[Nolano.ai @NolanoOrg]

🚀 SpectraSuite of Ternary and FP16 LLMs 🚀

We’re thrilled to release the Spectra Suite of open ternary (TriLMs) and FP16 (FloatLMs) language models from 99M to 3.9B parameters. At billion+ parameter scale, TriLMs upto 10x smaller can match the performance of FloatLMs.

1/5

[Apoorve Singhal @sapoorve]

Sharing my passion project: Swachh Parliament Chatbot! 🚀

Our chatbot allows you to enter your pincode and compare netas in Lok Sabha 2024 elections based on education, number of criminal cases, party affiliation, and age.

Link in next tweet:

[Nikhil Shah @iamshnik]

@delhivery just keeps on disappointing, still no contact. They don't reply to customer support emails or chats. Have just hogged my items at their warehouse. Even after commenting on my last post, they didn't bother to connect w me. Just a horrible experience all around.

[Nikhil Shah @iamshnik]

What a trashy experience with @delhivery , shipped a product from my native place to Bangalore, it hasn't been delivered in 15 days. They've just kept it at their warehouse in Bangalore. Their support is even worse, no response on email, chat, and calls.

[Apoorve Singhal @sapoorve]

lately haven't been active on here ...

have been heads-down working on v1 of a tool that formats/beautifies powerpoint slides
(not for creating slides, only for redesigning them)

formatted slides are in powerpoint itself, so 100% editable

input & 3 (of 20) output slides shown

[Nikhil Shah @iamshnik]

Sick drop! Made with mehnat and AI

[Tejas Vaidhya @imtejas13]

Dropping my first EP:
Recap-2023. (Powered by @suno_ai_ )

This EP serves as a recap of my 2023. Stay tuned for the release of improved music masters and extended song versions coming soon.
1/6

[Nolano.ai @NolanoOrg]

We are pleased to introduce Hi-NOLIN, the best performing 9B Hindi-English Bilingual LLM.
Blog: https://t.co/RdzHBJ4yTb

[Nolano.ai @NolanoOrg]

🚀 Today, we released the LoRDCoder models based on our novel method for compressing code LLMs that can be combined with quantization and pruning.

https://t.co/gNY96ttqZy

[Nikhil Shah @iamshnik]

Wuhooo! Was waiting to try out this for a while, now I only need to get a mac 😂

[Tejas Vaidhya @imtejas13]

@NolanoOrg is excited to introduce Cformers 🚀 - a project that enables fast inference of state-of-the-art AI models on your CPU!
But that's not all - Cformers also automatically downloads and caches these models for optimal performance.
Github Repo: https://t.co/m465pe6ZxG

[Nikhil Shah @iamshnik]

Google pay rewards have become so useless, I prefer, "Better luck next time" over them

[Nikhil Shah @iamshnik]

Wow! That looks exciting, will lead to so many on-premise AI applications 🔥

[Nolano.ai @NolanoOrg]

🚀 You can now achieve GPT-3 level performance on your Mac at 12 tokens/sec using compressed LLaMa 7B and optimized inference with just 4GB of RAM. Join our Discord for more updates: https://t.co/n6O0PdpH0b.

#GPT3 #ChatGPT #AGI #LLaMa

[Tejas Vaidhya @imtejas13]

Our @NolanoOrg’s SayIt(talk with the webpage) did this before well funded companies.
cc: @_AyushKaushal

[Nikhil Shah @iamshnik]

Ooof 🔥

[Greta Thunberg @GretaThunberg]

this is what happens when you don’t recycle your pizza boxes