Replit CEO Outlines Self-Driving Company Built on Agents
Replit CEO and co-founder Amjad Masad says the future of work is the "self-driving company," where AI agents do much of the grunt work while people are valued for judgement and taste. Speaking at TechCrunch Disrupt in San Francisco, Masad described how Replit has reorganized its own operations around that model. TechCrunch
Masad said Replit entirely changed how it operates, calling the closest analogy a "self-driving company" where a lot of work is driven by agents. The phrasing is deliberate. It does not mean autonomy for its own sake. It means agents carry tickets, code context, and operational state forward far enough that human time shifts to decisions only humans can make.
Support is the clearest reference implementation. Masad said Replit support tickets are first examined by an AI agent that performs initial triage using GitHub issues, the codebase, and Linear tickets. That grounds the response in the same artifacts engineering uses. No separate knowledge base to drift. No copy-paste from stale runbooks.
When a support staffer decides to submit a patch, it is "just a click to deploy another agent." The workflow collapses triage, code search, and fix generation into one continuous loop. A human still authorizes the action. The agent does the traversal across issues, code, and task tracking that used to consume the afternoon.
Masad said Replit staff now spend more time on discussions, whiteboarding, and architecture to figure out what to work on. That is a short sentence with large implications. Roadmapping replaces ticket-chasing. Specification replaces scavenging for context.
How Replit defines the model
Replit has laid out the idea in more detail outside the Disrupt stage. The company published a first-party post titled "The Self-Driving Company," and Masad expanded on the thesis in a separate write-up. Replit
The definition carries two constraints worth noting. Replit states a self-driving company is not one without people and that people still choose the destination. It separately describes a self-driving company as running on agents that act without a human re-checking every number. Replit
Put together, the model is not lights-out automation. It is delegated execution with human direction. Agents investigate production incidents, review pull requests, answer questions, and analyze business data, according to Masad's description of the internal deployment. The human role moves up the stack to intent, prioritization, and acceptance criteria.
Masad has also used blunter operational language for what management becomes in that structure, describing a CEO as a "glorified router" and saying the self-driving company thesis is where the future lies for Replit. He has linked that thesis to a larger enterprise ambition, saying Replit can create the operating system for the enterprise.
Enterprise context
Replit was founded in 2016 by Amjad Masad, Faris Masad and Haya Odeh and is based in San Francisco. The company describes itself as the world's leading online programming environment. It reached a $3 billion valuation on $150 million in annualized revenue last year. TechCrunch
Visa invested in Replit to power agentic payments for developers, a reminder that agent execution quickly touches authorization, money movement, and audit trails once it leaves the IDE. TechCrunch
For infrastructure-minded readers, the interesting part is not that agents can draft code. It is that Replit is using them as the connective tissue between systems of record. GitHub issues for history. Codebase for ground truth. Linear for ownership and state. Support queue for intake. Pull requests and production telemetry for verification. Each handoff that used to require human glue becomes an agent call with context attached.
The broader context here is familiar to anyone who lived through the shift from on-premises ticketing to chatops to platform engineering. Every generation reduces the cost of moving context between tools, and work reorganizes around the new bottleneck. In this author's view, Masad is correctly identifying that bottleneck as judgment. When triage and first-draft fixes are cheap, the scarce resource is knowing what is worth fixing, what good looks like, and what risk is acceptable.
Worth flagging for engineering leaders is the trust boundary Replit is drawing. Agents act without re-checking every number, but humans still choose the destination and still click to deploy the fixing agent. That keeps a human in the authorization path while removing the human from the search path. It is a pragmatic split. Full autonomy breaks auditability. Full manual review breaks throughput.
Looking at what this means for teams adopting similar patterns, the Replit example suggests starting where context is already structured. Support to code to ticket is a tight loop with clear inputs and observable outputs. Incident investigation, pull request review, and internal data analysis follow the same shape. Whiteboarding and architecture get the reclaimed hours, which is exactly where senior engineers add the most leverage anyway.
If the bet holds, the long-term payoff is less about headcount and more about cycle time. Small teams can carry larger operational surface area because agents carry the state between systems. That has, on balance, been the direction of good tooling for forty years, from compilers to version control to CI. Agents are the next layer for carrying routine cognitive load, leaving people to do what tooling has never done well, decide what matters.
