Divyakush Punjabi is a full stack and AI systems engineer based in Vellore, Tamil Nadu, India. He led engineering at LMX Labs across two live platforms, reads a B.Tech at Vellore Institute of Technology and holds a Major Degree in Artificial Intelligence from IIT Ropar, and ships production systems — multi-tenant SaaS, semantic retrieval, and machine learning services that run.
Selected work: Saturdays, DineGuru, GovernAI Research Atlas, GovernAI Studio, Content Recommendation Engine, AI & IoT Rockfall Prediction.
Who is Divyakush Punjabi?
Divyakush Punjabi — known simply as Divyakush, and romanised Divya Kush Punjabi where the name is written as two words — is a full stack and AI systems engineer based in Vellore, Tamil Nadu, India. He led engineering at LMX Labs as Software Engineering Lead, where he shipped two platforms that run in production, and built the retrieval side of GovernAI’s open-source research tooling.
He reads a B.Tech in Computer Engineering at Vellore Institute of Technology and holds a Major Degree in Artificial Intelligence from IIT Ropar, read concurrently and conferred in June 2026.
What does Divyakush Punjabi build?
Production systems, in three areas: multi-tenant SaaS, semantic retrieval, and machine learning services that are actually served rather than left in a notebook.
Four of them are live products. Saturdays is a consumer food-delivery platform with discovery, ordering and PhonePe and Stripe payments. DineGuru is multi-tenant restaurant SaaS covering inventory, kitchen ticketing, recipe costing, procurement and billing analytics on one core. GovernAI Research Atlas is a semantic search unifying papers, repositories and governance resources over a ChromaDB vector index. GovernAI Studio is an AI-governance training simulator built on Django, Celery and a hybrid RAG inference pipeline.
Where is Divyakush Punjabi based?
Vellore, Tamil Nadu, India — the campus of Vellore Institute of Technology, where he reads his B.Tech. He works to GMT+5:30.
The record on this site runs across four Indian cities: Vellore and the VIT campus in Tamil Nadu, Rupnagar in Punjab for the IIT Ropar programme and its convocation, Surat in Gujarat, and the IIT Bombay campus in Mumbai for the National Entrepreneurship Challenge. Engagements have been run both on-site and remote.
Where does Divyakush Punjabi study?
Two degrees, read in parallel. The core is a B.Tech in Computer Engineering at Vellore Institute of Technology, Vellore, Tamil Nadu, running 2024 to 2028, currently at a GPA of 8.17 out of 10.
Alongside it he read a Major Degree in Artificial Intelligence at the Indian Institute of Technology Ropar, Rupnagar, Punjab, from January 2025 to June 2026 — machine learning core through to a capstone: a transformer-based sequential recommender at 98.47% AUC-ROC.
What is Divyakush Punjabi’s professional experience?
5 roles across product startups, an AI governance lab, and a BASF partner.
Software Engineering Lead at LMX Labs, sole technical lead across two live platforms and the team building them. Web Developer Intern at GovernAI, shipping three products in one internship. Web Developer at VUBS Corporation, sole engineer on a full-cycle corporate build that cut Time-to-Interactive by 40%. UI/UX Designer and Frontend Intern at LayOver, where the mobile flows he researched and rebuilt lifted retention 20%. And Team Lead of Visionary Ventures, the team representing VIT Vellore at E-Cell IIT Bombay’s National Entrepreneurship Challenge, AIR 140.
What technologies does Divyakush Punjabi work with?
On the server: Python, FastAPI, Django, Celery and PostgreSQL, with ChromaDB for vector retrieval. On the client: React, TypeScript, Next.js and Tailwind CSS. For machine learning: TensorFlow, Sentence-Transformers and retrieval-augmented generation.
Closer to the metal, the work runs to Verilog HDL, embedded systems and computer vision at the edge — an adaptive traffic controller in RTL, and Netra, a closed perception-to-motion loop over MQTT.
Is Divyakush Punjabi available for work?
Yes — for full-time software engineering and AI roles, and for contract work, across India and remotely. Both of the 2026 engagements finished in July 2026.
Previous engagements have been run on-site, remote and on contract, so either arrangement is familiar rather than an experiment.
How do you contact Divyakush Punjabi?
Through the contact form on this site, which is the only route in — the address and phone number are deliberately not published. It reaches him directly.
He is also on LinkedIn at linkedin.com/in/divyakush-punjabi, on GitHub at github.com/Divyakush2006, and writes at dev.to/divyakush.
Divyakush Punjabi