i love writing!! it gives me an excuse to learn topics and is my small way of giving back to the community (inspired by Andrej Karpathy's videos). my past blogs: 1. a blog on lcms (10k readers) 2. llama3 from scratch (14k github stars, ~100k readers) 3. an rl guide (20k readers)

bootstrapped my company’s account with it’s first 1k - with the intention of building great memories, iterating towards great products & eventually arriving at a 1,000,000x multiplier.

my life right now - talking to users, shipping features, going on walks. on loop, every single day, 12-15 hours a day. my beta users are now using the platform on their own & not because i’m forcing them to use it. lowkey excited for the launch!!

it's been 2 weeks of exploring ideas & 5 weeks of building out a product. my beta users seem happy, i'm happy, it's time to ship it. i'm launching in 2 days, added 10k in liquidity for gpu rentals, hopefully y'all like it. going to take a break today & enjoy the weather!!

today, i'm excited to release cloudy -- a platform which enables you to sync & mount storage volumes on gpus sourced from popular cloud providers like lambdalabs, runpod, hyperbolic, nebius etc. save your team countless hours every month.

if you are going to run experiments on these models, rent your gpus on cloudy.so!! i’m trying to build the cheapest, prettiest, most convenient gpu rental product. i’m here to support you, ask me / dm me questions you have!!

in the next few weeks, i'm going to build the world's most convenient tools for training models, from fast throughput 500TB multi-region network volumes, short-term gpu nodes, llm/image eval tools, huggingface dataset/model caching, and so much more.

i’m currently bootstrapped & solo, which means you can essentially hire me to handle your gpu training infrastructure for FREE by signing up to try cloudy’s new product. accepting 5 serious companies, currently at 2. i can move extremely fast! dm me, thanks :)

if you are a founder, it’s worth spending 4-6 months on the product without raising money. with so much money available right now, it can be tempting but this delay benefits you in the long run. people that raise too early, sacrifice way too much to save on a few months on rent.

i spent the last few weeks carefully building compute, petabyte scale storage and caching infrastructure that let’s people switch between different gpu configs, compute providers & regions seamlessly within a few seconds. i’m launching this weekend & genuinely really excited :’)

Announcing Cloudy, a platform that seamlessly handles your training infrastructure. You can rent a single H100 or a cluster of 1000 H100s, manage petabyte-scale storage volumes & seamlessly go from running experiments to managing large scale training runs on a single interface.

In the next few years, we are going to see an explosion of teams training models, researching ideas or running experiments on GPUs. The GPU shortage problem will get worse, while the number of teams renting compute will increase. Cloudy is betting on this :)

my goal @ cloudy: onboard 100 startups by the end of the year. i'm extremely focused on building the best compute product in the world. it's going to happen, slowly and then all at once :)

i spent the last few days talking to startups, demoing cloudy, gathering feedback. made good progress on the goal, my makeshift crm dashboard is happy :) on sunday, cloudy will be the best compute marketplace in the world. it’s coming together & I AM LOCKED IN.

everyday, i wake up & get reminded that working on cloudy is the most self fulfilling and important project i could be working on. it took me years to find this level of conviction on anything i’ve ever done. i’m happy & focused :)

Introducing Cloudy, an end-to-end platform for training/running experiments on models. Spin up GPU infrastructure, manage storage, run benchmarks, all in one place. > Why Cloudy? Teams renting compute are often forced to use multiple cloud providers. This can be extremely

for a while now, i’ve been experimenting with different daily schedules. my most productive days are: 6 hours of focused work -> 4 hours break (beach, gym, lunch, walk, think) -> 6 hours of focused work. the break in the middle makes 12 hour days, 7 days a week very sustainable.

i just made cloudy’s account, it has 1 follow at the moment (me) 🥀, if you are interested please follow cloudy thanks!!

2026, ????

as the year comes to an end, it’s becoming increasingly obvious that 2026 is the year of creativity. stay humble, stay creative, go crazy!!! happy new year :)