
full stack / ai engg @ stealth ✧ builder @nebiusai ✧ guild @FactoryAI ✧ cooking on how LLMs sees your product. ✧ dm for work.
ping me here -
Joined June 2021
- Tweets6.7K
- Following194
- Followers362
- Likes11.4K
Pinned Tweet
building Tax Mitra for @waitin4agi_'s #BuildWhatMovesIndia hackathon!
the idea is to make income-tax notices actually understandable.
instead of pages of technical language, Tax Mitra turns the notice into clear questions, a personalized checklist, and an editable response while keeping the taxpayer in control.
your tax notice, made understandable.
google is used by almost everyone, from individuals to businesses.
and now they're letting openai into google cloud. that's a very interesting move.
there was speculation that google will drop a banger model by the end of this year, but after seeing this, i'm not sure anymore
OpenAI models are coming to Google Cloud.
We're bringing our frontier models and Codex agentic workflows to Google Cloud's Model Garden, so enterprises can use our technology where they already build.
openai.com/index/openai-mode…
this is the only design skill which you need man.
i have tried bunch of different design skills, but this one is really great, have been using it for all of my recent builds and it's surprisingly good.
you just need a proper design reference, rest all the things can be done
is this gonna be the next whop?
The wait is over.
Temple Batch 01 pre orders open November 1. Each @temple is hand-made to order and delivered to you within three months of your purchase.
i feel like a lot of companies will start building their own “AI labs” to handle internal workflows and automation.
but if Dark is specifically training models for Eternal’s products, that could be a game changer for them.
but if it is for consumers, I just hope that it should be something genuinely innovative and not another basic AI wrapper :)
finally, i'm part of the @FactoryAI Guild!
expect more builds coming your way.
we're gonna be cooking 🔥
thank you @adityawaslost for all the help, bro!
I’ve added something new to my daily routine. After work, I’ve started watching web design breakdowns and learning how to design an entire brand.
I used to be a web designer, so I’m brushing up on my skills. But more than that, it’s helping me develop the design taste I need.
let’s be honest, in software, people do judge a product by its cover. the better the design and UI/UX, the more enjoyable and intuitive the product becomes to use.
the scariest part about shipping a product isn't the bugs you know about.
it's the vulnerabilities you don't even know exist.
meet ShipSafe
55 security checks. real evidence, zero AI hallucinations, and results in under 30 seconds.
built with Droid by @FactoryAI and microVM from instabox.sh
link: shipsafe.runs-on.dev/
if you think that deploying is a headache, @deploywithox is built to solve that.
get your project deployed in 3 steps.
check it out!
Putting websites on internet shouldn't be complicated.
Introducing Ox ( @deploywithox ) . i have optimised it's UI/UX/DX for the lowest cognitive load.
1. Connect GitHub and your vps.
2. generate ox.toml config.
3. push to deploy.
the only thing which I get from my Google AI Pro plan is the 5TB storage that's it.
will try this now, atleast would love to experiment with that 80 gigs of VRAM.
Google AI Pro plans now package premium Colab benefits, landing you A100s with 80GB VRAM (Ultra subs even unlock H100s 🔥).
Here’s what that 80GB headroom unlocks directly in your browser:
✨ Run Gemma 4 31B in bf16
✨ Full fine-tune up to Gemma 4 E4B in bf16
✨ LoRA tune up to Gemma 4 26B A4B in bf16
✨ QLoRA tune up to Gemma 4 31B in bf16
pov you bought codex sub at the worst time
got some beefier GPUs
so here's free DeepSeek 4.1 Flash uncensored
~41m tpm shared
~403 tps peak
use it for cybersec, coding, whatever
at platform.xplabs.ai
Readers added context they thought people might want to know
The free models promoted in this post, such as DeepSeek V4.1 Flash, come from Experiential Labs which discloses inserting contextual sponsored content into responses to fund free or discounted access. Users report ads appearing directly in outputs.
platform.experientiallabs.ai/privacy
x.com/chaotictransfe…
x.com/DreamwovenArts…
platform.experientiallabs.ai/models
bro the amount of free models which are there on opencode rn is insane
muse spark 1.3 free (1M)
deepseek v4.1 flash free
mimo v4 free
cat jam free
GLM 5.3 flash free
and now this, like just keep shipping man!
it's crazy how fast the industry is changing.
i'm working at a stealth startup where we're constantly testing new harnesses, models, skills and agentic workflows to get more work done.
meanwhile, my friends and i graduated 3 months ago, and they're still grinding DSA and asking me for referrals.
DSA still matters, but the game is changing fast.
knowing how to build, ship and work with AI agents is becoming just as important.
i genuinely don't want them to miss out on what's happening.
please @Cloudflare even I wanna have this :)






















