currently shipping flickstat · open to roles ’26
Arnav Gupta
Computer-vision engineer & full-stack builder, taking problems from the training loop to the interface someone actually uses.
A free, analytics-native football platform. Built solo on a 3-person team — ~1,200 visitors in month one with zero paid marketing.
One-click image labeling on Meta's Segment Anything Model with multi-format export. Cut manual annotation effort by ~80%.
Config-driven platform turning raw file/DB/API data into trained models — reproducible, versioned runs with a React control UI.
Shipped a SAM auto-annotation tool, an LLM model-eval report generator, and a retraining-free aerial tree-detection pipeline.
Built a YOLOv8 culvert-detection pipeline for drone imagery; tuned thresholds and expanded the dataset.
Ran ops for a founder meetup, coordinated a 4-person team, and secured a free Andheri venue via a DevX co-marketing partnership — removing the event's biggest cost.
I spend a lot of my time inside India's tech community — not just attending, but hosting. I run and show up at Mumbai tech meetups, connecting builders and turning rooms full of strangers into people who ship together afterwards.
Hackathons are where a lot of my projects start — tight deadlines, a blank repo, and a working demo by the end. It's the fastest way I know to learn a new stack and to find the people worth building with. That same instinct led to my Startup School run, owning on-ground ops for a founder meetup and pitching a co-marketing partnership that landed a free venue.
Got a role, a project, or just want to nerd out about computer vision? My inbox is open.