Manmohan Singh Charan · product builder
1store
I · Blinkit offline · 2019–2020
I wired 450 neighbourhood stores across Delhi NCR into one network.
Early PM (APM title), offline team
Scale one: a city. 100 stores in mid-2019, 300 by September (press), about 450 by Oct 2020 when I left (my count). All in Delhi NCR; where each store sits is approximate.
Loading
1person
II · Tago · 2021–2022
My own bet: people should own their health records, in a form machines can read.
Co-founder & CPO
Papers from many clinics, filed into one record a person owns, structured so software can read it. Live on Google Play in 2021, on India's national digital health stack (NDHM).
Loading
1shopkeeper
III · Badho · 2022–2026
I co-founded Badho: shops buying stock straight from distributors, with data on two million of them.
Co-founder & Head of Product · B2B commerce for India's neighbourhood shops
Scale two: the country. Data on 20 lakh retailers across 11 states; 1M+ installs on Google Play. The places are approximate.
Loading
23:41
1orders a day
IV · GobbleCube · 2026–now
Every night, brands set tomorrow's discounts for the whole country from an engine I built.
Lead PM, Promo Planning
Scale three: the country, every night. Dots are shoppers across ~9.5M daily quick-commerce orders (Datum, Aug 2026). Gold marks the 400+ cities where tomorrow's discounts land. Discount figures are illustrative.
Loading
V · now
Now for machines
Four times I made a messy world legible for people. Now I'm doing it for models: giving agents real jobs, clear scores and places to practise. I build with agents every day, with seventy-plus skills of my own. This site was made that way.
Agents, harnesses, evals
What I'm building next
Work in progress on my own machine. Not shipped products.
Looking for remote roles where models meet the messy world.
Counts are exact. Places on the maps are approximate. India's map follows the Survey of India boundary.Built with agents · 2026