> ships AI-native products end to end: model pipelines, mobile, web, payments, growth.
16 years running national supply chains for Samsung, Haier, Tata Croma and Eureka Forbes taught me where AI pays off. Now I design, build and launch products where the AI does real work: a food-photo nutrition coach on Google Play, an autonomous YouTube studio that writes, voices and edits its own videos, an Arabic CV writer, and an Islamic calendar app used daily. Every project below is live or in flight, and runs on infrastructure I set up myself.
By day I run national spare-parts planning for Eureka Forbes as Head of Planning, Spares. My career has been about keeping service networks supplied at scale, through forecasting, S&OP, warehousing and sourcing for India's biggest appliance brands.
That is where my AI work started. I used machine-learning forecasts and automation workflows on real supply chains first, then began shipping my own AI products under Coderbydream.
3,000+ SKUs, 19+ warehouses, 700+ service centres, team of 10. Built an AI workflow that flags delayed orders and triggers corrective action.
Ran analytics for the service and warranty business. Before that, set up Croma's national spares network for multi-brand warranty plans.
Pan-India spares: 8 warehouses, ₹50 Cr inventory, 10,000 SKUs, 600+ franchises. Statistical forecasting lifted planning accuracy to 92%.
Material planning and supplier forecasts. Selected twice for supply operations workshops at Samsung HQ, South Korea.
LLM pipelines with validation and retry loops, vision models for food recognition, image and voice generation, cost tracking down to fractions of a cent per call.
Cross-platform apps and web products with auth, i18n in six languages, offline-friendly PWAs, PDF generation and subscription billing.
Serverless backends, edge databases, CI that builds signed Android bundles and deploys on push, cron-driven content pipelines with human approval over Telegram.
$ tower status --all