I build production AI systems where something real is at stake — tight budgets, no internet, live phone calls. Constraints that make the standard solution impossible are where I do my best work.
| Achievement | Impact | |
|---|---|---|
| 💰 | 94% AI cost reduction | 5,00,000+ students served |
| 🤖 | ₹30L+ saved annually | Multi-agent pipelines replacing manual workflows |
| 📞 | Voice AI under 500ms | Live telephony · custom Cerebras CS-2 architecture |
| 📱 | Full LLM offline in 1 GB | Quantised on-device model · zero cloud dependency |
🎬 vid2animVideo → 3D Character Animation Convert any video into production-ready 3D animation using AI pose estimation and universal skeleton retargeting. Supports Mixamo, VRM, Blender Rigify, Unreal, and more.
|
🎙️ AuroraReal-Time Voice AI Telephony Custom 3-stage inference architecture beating standard LLM APIs by 5–10×. Under 500ms end-to-end on live phone calls. 99.7% production success rate.
|
|
npm Package · AI Drawing Engine
Canvas engine giving AI agents a programmatic drawing hand. UUID-keyed shapes, 13 ES modules, 111 weekly downloads at peak.
|
Medium · Production AI Insights Writing about productionizing AI, solving edge cases, and moving beyond GenAI wrappers. |
Auto-fetches from my Medium profile — updates every time I publish.
📖 Authored a chapter in a published book on Artificial Intelligence
📜 Filed 6 copyrights · 1 patent application in process
🏆 Won national-level AI project competitions 3×
👥 Mentored 20+ students on AI implementation
Open to new opportunities and consulting. Heavy optimization, architecture bottlenecks, multi-agent systems, AI cost reduction — let's talk.
🌐 mehulcode12.github.io · 💼 LinkedIn · ✉️ [email protected]
The problems I enjoy most are the ones where the standard solution doesn't exist yet.