Vibhav Aggarwal
New Delhi IST
WorkResearchCodeAboutContact
← All work
2024.7Infrastructure

AI engineering

Where the descent began. I set out to learn AI engineering — and discovered I couldn't really do it without owning the layers beneath.

→Calling an API → understanding what runs it.

Part of the weave

2025.5

Server engineering

To do AI on my own terms I had to run my own servers — so I learned to. One layer deeper.

2021

GPU rig → ML engine

The mining rig, repurposed. Once I understood the hardware, it stopped being a miner and became my machine-learning engine and the seed of a self-hosted cloud.

2026.2

Factory OS

Two WhatsApp-native LLM agents running a real manufacturing business end to end — one for office operations (vendors, payments, CRM), one for the factory floor (production, dispatch, machines). Their memory is synced live across servers, and a ticketing dashboard ties both into one operational picture. In production since March 2026.

← All workGet in touch →