Analysis
Meta acquired Swedish AI agent startup Stilla to accelerate development of Meta Business Agent, its system for helping merchants handle customer conversations and transactions across WhatsApp, Messenger and Instagram, Axios reported September 9. The deal comes eight months after Stockholm-based Stilla emerged from stealth with $5 million in pre-seed funding, ArcticStartup reported separately.
What Meta actually bought
Stilla was founded in 2024 by Siavash Ghorbani and Kaj Drobin, both former Shopify executives who previously worked together at Tictail before Shopify acquired that company in 2018. Stilla built a shared-context layer for teams and AI agents, tying into tools like Slack, GitHub, Linear and Notion so that people and agents could work from the same live memory rather than each guessing what happened in a prior meeting or thread -- infrastructure Meta plans to fold into Meta Business Agent, which has now crossed 1 million businesses using it to manage customer messaging and commerce.
Stilla says its platform will keep operating as a service for existing customers even as its team and technology join Meta -- a continuity commitment acquirers frequently make and just as frequently wind down once integration is complete, since the acquiring company's own roadmap, not the acquired product's existing customer base, ultimately determines how long a standalone product survives inside a much larger organization.
Talent and technology over revenue
At eight months old and $5 million in disclosed funding, Stilla is a small deal by 2026's megaround standards -- but it fits a pattern that has defined this year's Big Tech AI acquisitions more than headline-grabbing multibillion-dollar deals have: buying a small, technically strong team and a specific piece of infrastructure rather than a large revenue base. Pulse has tracked Meta's broader AI agent push throughout 2026 as the company races to commercialize agentic offerings that are costing it billions in compute and R&D spend, with Business Agent's transaction volume across WhatsApp commerce specifically identified as a near-term revenue opportunity.
The risk for Stilla's existing customers, and for any startup building similar shared-context infrastructure, is the same one every small AI-infrastructure acquisition carries: Meta's stated continuity commitment is not binding, and the products most likely to survive long-term inside Meta are the ones that map most directly onto Meta's own roadmap -- in this case, Business Agent -- not the ones serving Stilla's original, broader customer base.
