About
I'm Mukund Chopra. I've had two careers and they happened in sequence, not in parallel.
The first career was global corporate and venture. Queen's Commerce, HEC Paris for strategic management, then a run through Citibank, Tata Steel, Olam, and Millicom across multiple countries. I was senior leadership at Groupon through the IPO. I was employee number three at Granify, where I architected revenue from scratch. I ran MD HomeCall as CEO — Canada's largest physician home visits provider — until it was acquired. I led North America for Universal Avenue. I was an early investor in Slack through Spaces before the acquisition and the liquidity event. I became a venture partner at Panache Ventures and a principal at Chopra Ventures.
I've been a CXO of venture-backed startups that raised money and went nowhere. I've been inside two IPOs. I know what the inside of a scaled organization looks like. I also know what the inside of a startup that doesn't survive looks like. The pattern recognition from that decade is the foundation for everything I build now.
The second career started around 2017 when I founded what became HomeEasy. I looked at the apartment locating business — one of the largest offline brokerage categories in the US — and saw a system hiding inside a sales process. A realtor takes requirements, queries inventory, presents options, closes. That's an agent with a database. I built the machine to prove it.
Today HomeEasy processes 30,000 leads a day across multiple metro areas — Dallas, Houston, Austin, Denver, Chicago. Most of that volume is handled by AI systems I designed and built myself. We hit 97% contribution margins in an industry where 20% is considered good. The humans who remain in the operation exist for exactly two things: phone calls and apartment tours. Everything else is automated.
I didn't start as an engineer. I became one because the engineers I hired couldn't build what I saw. I taught myself to code, then machine learning, then — when LLMs became production-ready — I replaced the engineering team entirely and started building solo with AI as my collaborator. Not as a novelty. As the actual operating model.
For most of the last few years, I thought I was unemployable. Not in the self-deprecating way — in the genuine "I've been in a bunker building weird things and I have no idea if anyone values this" way. The volatility of building alone is real. It's not just financial. It's the doubt, the isolation, the stretches where the systems break at 2 AM and there's no one to call.
About three months ago, something shifted. I realized the skills I'd built — orchestrating AI agents at production scale, building autonomous systems that run overnight, collapsing entire business functions into code — aren't niche. They're exactly what every company is trying to figure out. The death of doubt wasn't sudden. It was looking at the architecture I'd built and realizing: this is what the future of operations looks like, and I've been living in it for years.
I've deployed versions of this technology into private equity portfolio companies — quoting engines, CRM systems built from email mining, AI-powered sales operations. The pattern is always the same: take a human-intensive process, decompose it into primitives, automate the primitives, keep the humans only where presence or judgment is irreplaceable.
agent-ic is not a company. It's not a fund. It's a place to think about what I've built, what it means, and what happens when the same architecture gets applied to other industries. After years of saying no to inbound inquiries about our technology, I'm open to conversation. I don't have a pitch deck. I have seven years of production systems that work.
I don't know where this leads. That's the point.