new developer sign ups on micro1.

total = 61k

.Sean Rad x micro1

Patrick Collison on the importance of beauty and craftsmanship when building products

“If Stripe is a monstrously successful business, but what we make isn’t beautiful, and Stripe doesn’t embody a culture of incredibly exacting craftsmanship, I’ll be much less happy. I think the…

huge breakthrough at micro1 this week! we released an AI bot that is saving us $20k/month with its first version.

we recently onboarded ~500 engineers with one of the largest AI companies & started receiving more than 10 queries an hour from developers. our developer success…

should u fine-tune, RAG, build foundational model, etc?

this diagram made by the AI team at micro1 answers this.

download & zoom in.

We onboarded 500 engineers with one of our clients and started getting more than 10 queries an hour from engineers.

Our developer success managers were overwhelmed. We created a dev success bot that now resolves more than 50% of queries instantly -- right within slack threads.…

Neuralink was kind enough to open their doors for me to tour their headquarters a few weeks back. It was an amazing experience and a day I'll never forget. This was at a company wide meeting at the end of the day. Thank you to everyone who made this possible. Hope y'all enjoy!…

async makes micro1 ~2x faster. i would never do it any other way.

30 min meetings become a thread on slack with ~5 well thought out messages instead.

for gpt-vetting proctoring, we now run the ML algorithm locally to calculate the trust score so we don’t have to store any sensitive video-based or screen sharing data even for a moment. privacy 1st always at micro1.

13k async technical interviews have happened on micro1

that's ~10k hours saved

that's 1 week of work for 250 engineers. these hours were spent on building product instead of interviews. most engineers hate interviewing anyways.

AI-powered async technical interviews will be…

micro1 is giving me 240 hours of free software dev work for my twitter followers

they made the z fellows website

more info here: micro1.ai/partners/zfell…

we're now using OpenCodeInterpreter-DS-6.7B for the coding exercise part of gpt-vetting, no longer gpt-4. performing much better & running a fine-tuned version of it on a single GPU.

reality is, OpenAI will quickly lose market share with open source models that have much less…

goal by end of 2024: replace the entire micro1 interview process with gpt-vetting. and replace DDP creation completely with robust gpt-vetting report data. at that point, existing team of recruiters & interviewers will be the operators of this AI recruitment engine and we won’t…