AI automation startup Relay shuts down, staff joins Google's Chrome team

AI automation startup Relay has shut down after five years, with founder and CEO Jacob Bank announced on August 17, 2026 as Google's new VP of Product for Chrome — a move that multiple reports describe as a classic Silicon Valley acqui-hire dressed up as a strategic talent bet on the browser's future.

The Claim

Relay launched in 2021 with a pitch that positioned it as a more approachable, AI-native alternative to established workflow automation tools. Where competitors required users to wrestle with APIs and JSON configurations from day one, Relay emphasized a clean interface, AI embedded directly into automation steps, human-in-the-loop approval gates, and genuine multi-user collaboration rather than solo-builder tooling. The platform connected widely used business applications — Gmail, HubSpot, Slack, Notion, and hundreds more — and found a niche automating practical operational work: lead qualification, customer onboarding sequences, meeting follow-ups, and CRM updates.

The shutdown is being staged to give customers time to migrate. Free-tier users lost access on August 15; paying customers have until September 14, 2026. Relay has provided export tools for workflows and data, though every affected team still faces the work of finding and rebuilding on a new platform.

The headline news is Bank himself, not the closure. A former Google product manager who led Gmail, Google Calendar, and Google Chat, Bank previously sold his calendar productivity startup Timeful to Google in 2015. His return to the company — this time as VP of Product for Chrome — comes with a quoted rationale that reveals more about Google's intentions than about Relay's fate. Bank described Chrome as "a perfect place to collaborate with agents," signaling that his mandate is not simply to iterate on the browser but to transform it into a native execution environment for AI agents.

What We See

The framing from Google and Bank is that this is a natural alignment of vision — a founder who spent years building AI-native automation finding the right platform to take that thinking to global scale. Chrome is not a niche product: Gemini, Google's AI assistant, recently crossed 1 billion users, and the browser is the surface through which most of that reach is deployed. Bringing in someone whose entire recent career centered on making AI automation feel human and approachable is a coherent move if the goal is agent orchestration inside the browser rather than just a chatbot sitting in a sidebar.

The Whalesbook analysis frames this plainly as an acqui-hire — a pattern where large technology companies prioritize experienced teams over purchasing entire businesses. That framing is accurate as far as it goes, but it undersells what Bank's specific background adds. This is not a generic talent pickup; it is a hire targeted at the exact intersection of workflow automation and product accessibility that Chrome would need to support AI agents running complex, multi-step tasks across web applications. Bank's prior work building human-in-the-loop approval mechanics into Relay is the kind of domain knowledge that takes years to develop and cannot easily be replicated by retraining existing browser engineers.

Our read is that the Chrome agent narrative is the more consequential story here. Google has spent the past two years positioning Gemini as an assistant layered on top of existing Chrome behavior. A VP of Product hire with Bank's specific background suggests the next phase is something more structural: agentic capabilities that don't just answer questions but execute workflows across tabs and services, potentially with native permission and approval models built into the browser itself. Developers building automation tools — including anyone relying on browser automation stacks that depend on fragile lower-level interfaces — should watch this closely. If Chrome exposes higher-level, stable agent APIs, the tooling landscape for web automation changes materially.

Where It Falls Short

What the available reporting does not answer is what, specifically, Bank and his team are joining to build. "A perfect place to collaborate with agents" is an evocative phrase, not a product specification. The Mezha source that picked up this story added minimal new detail beyond confirming the basic facts of the shutdown and hire. The LinkedIn commentary from the automation practitioner community focuses more on migration advice than on the Chrome angle — understandable from an operator's perspective, since when your automation platform shuts down with 30 to 60 days' notice, philosophical questions about browser agent APIs are not the immediate priority. But it means the technical substance of Bank's mandate at Google remains unverified. Whether this leads to publicly available agent APIs, tighter Gemini integration, or something still in early design is genuinely unknown.

There is also a pointed irony worth naming directly. Relay was founded by a team that had previously built flagship Google products — Gmail, Calendar, Google Chat — and then left to compete with third-party automation tools that connected those very products. Five years later, the team is returning to Google, which will now internalize whatever product instincts Relay developed for making AI automation feel approachable to non-technical users. Independent automation vendors compete not just against each other but against the platforms they integrate with, and this shutdown is a clean illustration of that structural disadvantage. A well-regarded product, built by credentialed founders, serving a genuine user need, did not survive proximity to a platform player that can ship comparable functionality to a billion users at no marginal cost.

For developers currently running Relay workflows, the immediate concern is practical. Multiple sources confirm that export tools are available, but migrating automation logic between platforms is rarely as clean as exporting a file and importing it elsewhere. The LinkedIn practitioner community has surfaced n8n as one candidate migration path — an open-source self-hosted alternative that gives teams direct control over their automation infrastructure — though any evaluation should account for the operational overhead that running your own orchestration layer introduces.

Sources