Knuckle Up with Nakul

Parag Agrawal: "I only like jobs I'm slightly unqualified for" — Twitter lessons and Parallel's agent-web bet

Key points

Key takeaways from a Knuckle Up with Nakul interview with Parallel Web Systems founder and CEO Parag Agrawal (November 2026):

Curiosity over ambition. Agrawal — gold medallist at the International Physics Olympiad at 17, top of his computer science class at IIT Bombay and a Stanford PhD — says he was never an ambitious child: he was happy-go-lucky, did maths and physics because he and an exceptional peer group found them fun, and never had role models or idols.

Why Twitter. Disillusioned with academia's slow, low-signal peer-review loop, he joined Twitter as a junior engineer on the ads team, where a small team grew from around 10 to hundreds of people, scaled from four to 4,000 computers in about 18 months and took revenue from almost nothing to a couple of billion dollars in two to three years.

Rising through the ranks. He became Twitter's first distinguished engineer, then CTO, then CEO within a decade by surrounding himself with more capable people and working problems from the outside in — he flew to India to hire IIT undergraduates with no precedent, printing offer letters on the day of marathon six-hour interview sessions.

The CEO's job. Once in charge he stopped shapeshifting to fit the company and instead adapted the company to who he was — personality, vision and mental model dictate how a business should be organised and hired, and with a five-year view the CEO has to transform the business and own themselves.

Running Twitter in "dual reality". Through the contested takeover he held two contradictory truths — committed to closing the deal while operating as if it would not close — and credits lawyers on almost every decision and strong leadership for Twitter continuing to operate and ship throughout the period.

The agent-web thesis. After Twitter he built agents as a hobby "when they didn't really work", and the obsession became Parallel: if agents work they will use the web 1,000x more than humans, and that web needs new technology and business models — search built for AI agents, with plans to compensate content owners so the web stays open.

Why not the incumbents. He argues neither Google nor the frontier labs will build search for everyone's agents, leaving room for an independent player like Parallel — backed by Kleiner Perkins and Sequoia — which has grown roughly 4x to about 120 people in under a year.

Founder at 40. Asked why he would start a company at 40, he replied: "40 is young. I only like jobs where I think that I am slightly unqualified for doing that."

Read more: Knuckle Up with Nakul (YouTube)

More top stories