The AI scheduling agent that runs your calendar for you - so two calendars can finally talk to each other.
There is a specific, low-grade misery in scheduling a meeting. You propose three times. The other person is free for none of them. You propose three more. A timezone gets miscounted. Four emails later you have agreed to a slot that neither of you loves. Kais Khimji, who spent six years as a partner at Sequoia Capital, found this genuinely strange - strange enough to leave venture capital and build a company about it.
His framing is disarmingly literal. "I have a time database - my calendar," he has said. "You have a time database - your calendar, and our databases just can't talk to each other." Put that way, the back-and-forth email is not a social nicety. It is a missing API. For twenty years the two most important records of your time refused to speak, and humans were left doing the integration by hand, one awkward reply-all at a time.
Blockit's answer is an AI agent that you cc on an email or message in Slack. It reads the request, checks your calendar, understands which of your meetings are movable and which are sacred, and books the thing. When the person on the other end is also a Blockit user, something stranger happens: the two agents access both calendars and negotiate a time directly, with no human in the middle at all. The company says it has done this more than 100,000 times.
The idea is not new to Khimji. He first sketched it as a student at Harvard roughly a decade ago and shelved it, because the technology to make it trustworthy did not exist yet. What changed was not the calendar. It was the model underneath - large language models good enough to read the tone of an email, weigh a vague "let's find time next week," and act on it without embarrassing you. Some ideas, it turns out, are not wrong. They are early.
Co-founder Yoon-Suk "John" Han is the other half of the bet, and arguably the more improbable one. He has spent a decade working inside calendar products - Timeful, then Google Calendar, then Clockwise. If there is a small, unglamorous priesthood devoted to the problem of when-should-we-meet, Han is a senior member of it. Blockit is what happens when a person who understands the plumbing meets a person who understands why the plumbing matters.
"The calendar is the last untouched social network - and it can only now be unlocked with AI."
You do not open an app. You do not pick from a grid of times. You just include Blockit and let it do the coordinating - the way you would delegate to a very fast, very literal chief of staff.
cc Blockit on an email thread or message it in Slack. Plain language is fine.
It reads your calendars - multiple Google or Outlook - plus timezones, priorities and even email tone.
If the other party is on Blockit, the two agents settle a time between themselves. If not, it proposes and confirms.
The invite lands. Over time it learns which meetings you protect and which you let it move.
Stop trading availability over email. Blockit finds the slot and sends the invite while you keep working.
Multiple attendees, conflicting timezones, an in-person venue - the agent handles the fiddly parts.
Which meetings are non-negotiable, which can slide, when you like focus time. Behavior sharpens with use.
When both sides use Blockit, their calendars negotiate directly - the network effect the company is betting on.
Blockit's raise reads like a reunion tour of the people who watched Khimji's career up close - led by his former firm and joined by the former CEO of the company that basically invented the professional social graph.
| Round | Amount | Date |
|---|---|---|
| Seed | $5,000,000 | Jan 2026 |
Chart is illustrative - bars are scaled for reading, not to a shared axis.
Spent six years as a partner at Sequoia Capital, joining in 2019 and stepping back in 2024 to build Blockit - an idea he first had at Harvard a decade earlier.
A decade inside calendar products - Timeful, Google Calendar and Clockwise - now applying that hard-won intuition to autonomous scheduling.
Automated scheduling has a graveyard. x.ai and Clara Labs both tried to build AI meeting assistants in the mid-2010s and both shut down, in part because the models of that era could not be trusted to act unsupervised. Blockit's implicit argument is that the timing is finally right - that "zero humans in the loop" is now a feature you can say out loud rather than a liability you hide.
The giant in the room is Calendly, the booking-link company last valued around $3 billion. Blockit's pitch is not a better link. It is the claim that the link itself is a workaround - a way of making a human do the agent's job. If agents can negotiate directly, the scheduling page becomes a relic. That is a big if, and the history above is why it is worth watching closely rather than assuming.
Product walkthroughs and interviews live on the company's own channels and in press coverage of the launch. Start here.
Blockit AI is a San Francisco startup building an AI scheduling agent that runs your calendar for you. You cc it on an email or ping it in Slack, and it books, reschedules, and negotiates meetings autonomously - learning which meetings are movable and which are not. When two Blockit users need to meet, their agents talk to each other directly, skipping the human back-and-forth. Founded by former Sequoia partner Kais Khimji and calendar veteran John Han, the company raised a $5M seed round led by Sequoia in January 2026 and says it has coordinated more than 100,000 meetings across 200+ companies.
Last updated: