Former Sequoia partner’s new startup uses AI to negotiate your calendar for you

Kais Khimji has spent the bulk of his professional life on the investing side of startups, including six years as a partner at Sequoia Capital. But like several other former Sequoia partners — among them David Vélez, who went on to found Brazilian digital bank Nubank — Khimji has long wanted to build a company of his own.

On Thursday, Khimji announced that he has revived an idea he initially explored as a student at Harvard University roughly a decade ago, transforming it into an AI-powered scheduling startup called Blockit. In a notable show of confidence, Sequoia — Khimji’s former firm led Blockit’s $5 million seed round.

“Blockit has a chance to become a $1Bn+ revenue business, and Kais will make sure it gets there,” wrote Pat Grady, a general partner at Sequoia who led the investment, in a blog post announcing the round.

Automated scheduling is not a new problem space. Over the years, numerous startups have attempted to streamline calendar coordination, often with limited success. Khimji argues that recent advances in large language models now make it possible for Blockit’s AI agents to handle scheduling in a far more fluid and effective way than earlier efforts, including now-defunct companies such as Clara Labs and x.ai — the latter of which ultimately gave its domain name to Elon Musk’s AI venture.

Unlike Calendly, the current category leader that was last valued at $3 billion and relies on users sharing availability links, Blockit is betting on a more autonomous approach. The company believes its AI agents can manage the entire scheduling process end to end, without requiring humans to coordinate or manually confirm availability.

Khimji and co-founder John Han — who has previously worked on calendar-related products including Timeful, Google Calendar, and Clockwise — describe Blockit as something closer to an AI-driven social network for time.

“It always felt very odd,” Khimji said. “I have a time database — my calendar. You have a time database — your calendar — and our databases just can’t talk to each other.”

Blockit is designed to address that disconnect directly. When two people need to schedule a meeting, their respective AI agents communicate to negotiate a mutually suitable time, eliminating the familiar back-and-forth of email.

Users can activate the Blockit agent by copying it on an email thread or messaging it within Slack to request a meeting. From there, the AI handles logistics, negotiating both timing and location to align with each participant’s preferences.

According to Khimji, the system can operate much like a human executive assistant. Users provide guidance on how rigid or flexible their schedules are, specifying which commitments are fixed and which can be moved if needed. “Sometimes my calendar is crazy, so I need to skip lunch,” he said. “And the agent needs to know that it’s okay to skip lunch.”

Blockit’s agents can also be trained to weigh the importance of meetings based on contextual cues. For example, users can instruct the system to treat a message signed with a formal “Best regards” as more urgent than a casual request ending with “Cheers.”

By learning and applying these kinds of preferences, Blockit appears to be tapping into what venture firm Foundation Capital partners Jaya Gupta and Ashu Garg describe as “context graphs.” In a widely circulated essay, the pair outlined a multibillion-dollar opportunity for AI agents to capture the reasoning behind decisions by modelling the implicit logic that typically resides only in a person’s head.

Blockit is already in use at more than 200 companies, including AI startup Together.ai, the recently acquired fintech firm Brex, and robotics startup Rogo. Venture firms such as Andreessen Horowitz, Accel, and Index Ventures are also among its users.

The product offers a 30-day free trial. After that, Blockit costs $1,000 per year for individual users or $5,000 annually for a team license supporting multiple users, according to Khimji.