Over the past year I have been building an end to end pipeline for fine tuning, and in doing so, I accidentally came across a problem that apparently not many people are interested in solving, so diverged to build a solution I genuinely believe solves a real problem and now I'm at a fork in the road.
The problem we're solving:
Industries like automotive, healthcare, legal, and finance are sitting out of the AI wave not because they lack data, but because their data is too sensitive to use directly for fine-tuning. Nobody wants to feed real customer conversations into a training pipeline and hope for the best.
Iam building an app MaidCircle: Attendance & Salary Tracker which solves my use-case of tracking maid attendance & auto calculate salary. Sharing it here so that it could be helpful for others too.
Its currently in closed testing for Android & available in production for iOS.
Hi everyone, I m a judoka and I ve been practicing judo for about 10 years now. I m currently a brown belt and preparing for the different exams to pass my black belt. Over the years, I ve noticed something: outside of training sessions, it s actually hard to really analyze your own techniques. During practice, the coach gives feedback, but once you re home, it s mostly just memory and feeling. Sometimes I film myself, but it s difficult to objectively see what s wrong with my posture, balance, or timing. So I ve been thinking about building something. I m exploring the idea of an app that could analyze judo techniques from a video and give feedback on posture, angles, and positioning not to replace a coach, but to help judokas better understand their movements and track their progression over time. Right now, I m just validating the idea and trying to understand if other judokas would find this useful. Would something like this interest you?
What would you personally want such an app to focus on? I d really appreciate your honest thoughts.
https://judoai.vercel.app/
I've been using Claude Code daily for months, and while it's incredible, I kept running into the same frustration: I had no idea what was actually happening during sessions. Why did that refactoring cost $12? Why did it take 15 minutes? Is Claude reading the same config file over and over? The answers were technically there buried in .jsonl log files that I'd never realistically parse by hand. So I built Argus, a VSCode extension that just... reads those files and makes sense of them.
After 4 weeks of beta testing on TestFlight, I finally launched my app on the App Store 2 days ago. To my surprise, I hit 300 downloads in 48 hours, all from Reddit, without spending a dollar.
Ever wanted to ask "what did Cowen say about BTC dominance in 2023?" and get an instant, cited answer without scrubbing through hours of videos?
That's CryptoLens.
It ingests a crypto analyst's entire YouTube library into a vector database and lets you chat with years of historical analysis with source links and timestamps for every answer.
