YesPress Profile
Shishir
Mehrotra
"The man who sat on a document for four and a half years - and was right to."
CEO at Grammarly | Co-founder, Coda | Former VP, YouTube
"The man who sat on a document for four and a half years - and was right to."
He spent four and a half years building a product before showing it to the public. Not because he was cautious. Because he was certain.
In December 2024, when Grammarly announced it was acquiring Coda, the headline was about writing software swallowing document software. The real story was simpler: a founder who had spent a decade quietly rebuilding how teams think on paper got handed the keys to a platform used by 40 million people every day.
Shishir Mehrotra is not the kind of person who shows up in profiles because he wants to. He shows up because the things he builds eventually become impossible to ignore. Coda - the all-in-one work operating system he co-founded in 2014 - spread through the Fortune 100 the way all genuinely useful software does: because one person in a company used it, and then everyone else wanted to know what they were doing.
Before Coda, there was YouTube. Mehrotra joined in 2008 and became VP of Product, Engineering, AND UX - a title that almost never belongs to one person, because those three disciplines tend to fight each other. He ran all three for six years through the most consequential growth phase in internet video history. The platform he helped scale now serves over two billion users a month.
He is, in his own words: husband, father, maker, entrepreneur. In that order. This is either a carefully cultivated personal brand statement or an honest confession about where his head actually is. From the evidence of his public writing - 16 shared documents on Coda, including a Wordle habit tracker - it reads as the latter.
Do you have an idea you can't imagine not working on? Do you have a person you can't imagine not working with?- Shishir Mehrotra on his two criteria for starting a company
The earliest version of the Microsoft Word document appeared in 1983. The blank white rectangle with the blinking cursor. The toolbar running across the top. The Save icon that still looked like a floppy disk in 2010. For four decades, the document was not updated so much as it was ported - the same interface moved from desktop to browser to cloud.
Mehrotra noticed this at Microsoft. He spent six years inside three of the biggest productivity products ever made and walked out with a single insight: the document is the primitive unit of how organizations think, decide, and remember - and no one had actually reinvented it.
Coda's answer was to make documents programmable. A Coda document could contain a table that behaved like a database. A button that triggered an action. A view that filtered differently for different people. A form that wrote back to the same data that the table was reading from. Not a document and a spreadsheet and an app - just one surface that could be any of them, depending on what the team needed.
The pricing model reflected the philosophy. Coda charges only for "Doc Makers" - the people who create and configure. Viewers are free. This was not generosity; it was a bet that once a team's workflow lived inside a Coda document, the document would spread on its own. The bet paid off. 80% of the Fortune 100 used it. Most of them came in through a single team.
Before it became part of Grammarly, Coda had grown to become one of the most quietly influential pieces of enterprise software in Silicon Valley.
Mehrotra has become one of Silicon Valley's most-cited thinkers on team rituals and meeting design. His frameworks - all published publicly in Coda documents - have shaped how thousands of teams run their week.
When the Grammarly acquisition closed in December 2024, Mehrotra inherited something he had never run before: scale. Coda had 50,000 teams. Grammarly has 40 million daily active users. The operational difference between those two numbers is not linear.
The strategic opportunity is. Grammarly knows how people write. Coda knows how teams structure their thinking. Mehrotra's bet is that the combination - writing assistance embedded inside flexible, programmable documents, animated by AI - is the thing that comes after the productivity apps of the 2010s. Not a better Word. Not a smarter Google Doc. Something new.
In June 2024, six months before the acquisition closed, Coda launched Coda Brain: an AI assistant that let teams query their own company's knowledge rather than the internet's knowledge. Ask it who made a decision last quarter and it returns the meeting notes, the Slack thread, and the Coda doc where the decision was documented. It is the kind of tool that is obvious in retrospect and nearly impossible to build without a decade of understanding how teams store information.
Mehrotra does not talk about AI as a trend. He talks about it as a change in the medium. Documents that used to be read are now documents that can read themselves back to you. Decisions that used to be buried in email are now searchable by intent. The work of a team, previously stored in dozens of disconnected tools, can be made legible and queryable. This is the thesis he is building at Grammarly in 2025.
His self-description has not changed since he ran Coda: husband, father, maker, entrepreneur. He still uses Coda to track his Wordle results. He still publishes his frameworks publicly. He still answers the same two questions anyone who wants to start a company has to answer honestly. The scale around him has shifted. The orientation has not.