Built GitMarks, a Neovim plugin for sharing file marks across clones of the same repo.
Think of it as Harpoon, but shared.
Mark your important files, sync them to the cloud, and jump back to them instantly.
🎥 https://t.co/75BlhjBVnx
🔗 https://t.co/Bp1jnm2Ery
#Neovim#Go#Lua
Built Octogen 🚀
An AI-powered GitHub repository assistant.
I thought it'd take a weekend.
I was wrong 😅
Wrote about everything I learned while building it—from rate limits to vector search and production bugs.
📖 https://t.co/zdluaTlmeL
⭐ https://t.co/yg1c2WSxww
Most CV surveillance demos stop at object detection. Real security starts after that. I built EyeNet, a real-time system that turns noisy webcam feeds into actionable security incidents using tracking, event prioritization, and anomaly scoring. https://t.co/0somgYiOqJ
Built EyeNet 🎥 — a real-time campus surveillance system that turns raw video into structured, actionable incidents.
Face recognition + YOLOv8 hazard detection + ByteTrack + priority event bus, all the way to SMS/email alerts.
Wrote up the architecture 👇
https://t.co/0somgYiOqJ
I rebuilt a password manager — this time with the goal of creating something people could realistically trust, not just a demo.
Biggest lesson:
Auth is just the door, sessions and key management are the real security model.
Built with client-side encryption
#SecurityEngineering
We’re hiring 2–3 Software Engineering Interns at @MatiksHQ – a gamified learning platform making math fun and social.
What we’re looking for:
Experience with React / React Native
Some experience with Golang
Prior startup experience is a big plus
Curious mindset & love for building products
🗓 Duration: 6 months
💰 Stipend: ₹50,000 per month
Mode:- WFO (Whitefield, Bangalore)
🎯 Bonus: High chance of PPO based on performance
If you love startups, shipping fast, and learning by building – this is for you!
📩 Apply here: https://t.co/znYprfBMzv
Or DM me for details!
Good engineering isn’t always about making a feature work.
It’s also about understanding:
Performance, scaling impact of your code and how it is going to affect your users in the long run.
I recently worked on a PR to add curve support for custom shapes in a graph editor (Meshery/Kanvas) that uses Cytoscape.js.
What started as a feature quickly turned into a deep lesson on performance, rendering, and knowing when not to ship. 👇
PR = https://t.co/vY7BMg8Xk0
The realization:
Supporting curves by converting them into polygons is technically correct — but performance-hostile in Canvas-based graph engines.
This becomes a foot-gun when exposed to users.