I've shipped AI for users who don't forgive bad software: 7,000+ farmers, factory floors, and hundreds of senior professionals who were told the algorithm was not for them.
The teams who need AI most are not engineering teams. They are operations, sales, marketing, leadership — the people the tech industry has always expected to wait.
I do not make them wait.
Engineer, product manager, founder, farmer, AI builder — in that order.
I started in engineering — SSL-VPN internals at Juniper, computational biology at Strand Genomics, enterprise tools at Sun Microsystems. Then product at Amazon, where I shipped Omakase and Amazon Widgets — the latter became a multi-billion-dollar revenue driver. Then Yahoo, running the SMB platform and integrating the Lexity acquisition. In 2008, I co-founded Ohana Media — one of India's earliest real-time bidding platforms, bootstrapped to a $3M+ annual run-rate, exited via stake sale.
I could have stayed on that trajectory. I did not.
In 2017, I founded Farmizen — a farm-to-fork marketplace that scaled to over ₹10 crore in annual run-rate and connected 7,000+ farmers across two cities. BBC covered us. CNBC covered us. Bloomberg covered us. I spent nine years building technology for users with no laptops, patchy connectivity, and trust earned in person, over chai, in Kannada.
If a system needs its user to adapt to it, the system is wrong.
Then the platform shift of my career arrived, and I went all in. First at Dragonfruit AI Ventures, where I built and launched two ventures in a year: mantid.ai — real-time video intelligence for workplace safety on factory floors — and juvant.ai — AI agents for e-commerce operators. Production systems, not demos.
Now all of it feeds one thing: theCRUX — live cohorts where senior professionals learn to run AI agent teams. 600+ trained. Every cohort current to what Claude shipped that fortnight. Every graduate leaves with deployed work.
Underneath it all sits the Delegation Quotient — twenty years of watching technology adoption taught me the bottleneck is never the technology. With AI agents, the bottleneck is delegation. And it's measurable.
Computer Science from BITS Pilani, MBA from ISB — but the credentials matter less than the pattern. I keep choosing the harder problem over the bigger paycheque, and the users everyone else ignores.
Live cohorts where senior professionals — PMs, founders, CXOs — learn to manage AI agent teams. No code required, current to whatever Claude shipped that fortnight. Every graduate walks out with deployed systems. Also runs as dedicated corporate programs.
Explore thecrux.ai →I work directly with leadership teams on two problems: adopting AI in ways that ship results, not pilots — and building go-to-market systems that actually run. Every engagement starts with a 2-week diagnostic sprint: defined problem, defined deliverable, no retainer until there's a result on the table.
Book a consulting call →Two AI ventures built and launched in a year. mantid.ai — real-time video intelligence for workplace safety, with precision targets, latency SLOs, and fallback behaviours. juvant.ai — AI employees for e-commerce operators. A factory floor doesn't care about your demo.
India's largest organic farming community — 7,000+ farmers, ₹10 Cr+ ARR at peak, covered by BBC, CNBC and Bloomberg. Nine years building for users every algorithm ignored. The field where I learned that adoption, not capability, is the real frontier.
Explore Farmizen →Your IQ got you hired. Your EQ got you promoted. Your DQ — how well you brief, hand off to, and review AI agents — decides what happens next.
Everyone has the same models now. What separates 10× output from a slightly better search engine is the quality of delegation. DQ makes that quality measurable — twelve scenarios, three dimensions, one number.
Take the 4-minute assessment →I have spoken to corporate audiences, startup cohorts, and academic institutions. If you want someone who will say what they actually think — not what the audience wants to hear — I am available. Consulting runs as discovery sprints (2-3 weeks), implementation builds (6-12 weeks), or retained advisory.
AI-personalised pranayama guidance — custom routines based on your time, health, and goals. Built because I practise pranayama daily and every existing app felt generic.
pranaflow.app →Learn Sanskrit. Understand mantras. Built because I wanted to read the original texts — not someone else's translation of them.
devalingo.com →Both shipped solo, built with AI agent teams — the working method I teach. You no longer need a team or a funding round to put something real into the world.
BBC, CNBC, Bloomberg, and Indian media — for Farmizen's work in farm-to-fork commerce, agricultural technology, and startup building.
DQ: The Delegation Quotient — on the skill that decides who wins the AI decade. In progress.
I respond to specifics, not form letters. Tell me what you are working on, what you need, and why you think we should talk.
Book a Call →For everything else, use the form or email me directly.
shameekc@gmail.com