LocalStack, the Zurich startup known among developers for emulating the cloud on their own machines, has announced the acquisition of WonderTwin AI. According to the press release relayed on October 19 by Startupticker.ch, WonderTwin AI's emulators now make it possible to simulate SaaS applications locally as well, not just cloud infrastructure. The press release mentions neither the amount nor the terms of the transaction.

What LocalStack does

LocalStack offers a local cloud development platform that reproduces the behaviour of AWS and Snowflake on the developer's machine. Instead of deploying every change to a real cloud account to test it, the team works against a local copy that behaves like the original service. The company highlights faster and safer development, particularly for workflows in which AI agents write and ship code.

According to the press release, LocalStack has more than 1,500 customers and its Docker images have been downloaded hundreds of millions of times. The company describes itself as "the de facto standard" for local cloud development worldwide: this is its own claim, which the source does not allow us to verify.

The problem of third-party services in AI software development

A cloud application rarely depends on the cloud alone. It collects payments, syncs a CRM, measures product usage or relies on a code repository. To test these integrations, teams generally use real instances of the third-party applications, in shared "sandbox" environments.

According to the press release, these shared sandboxes cope poorly with the scale and concurrency of AI software development: when several developers and several agents are testing at the same time, the common environment becomes a bottleneck. The pressure shifts onto the teams, which are expected to deliver reliable applications at the pace that AI makes possible. The issue ties in with a broader observation about AI agents producing more code: speeding up one step shifts the load onto the next ones.

What WonderTwin AI brings

WonderTwin AI provides emulators of third-party applications, including GitHub, HubSpot, PostHog and Stripe, as well as 35 others, which run on local infrastructure. Each developer, and each AI agent, can thus launch their own sandbox to test and validate an integration without connecting to the real APIs of these services.

Together, the two companies want to offer on-demand sandboxes reserved for a single user ("single-tenant"), to validate the entire stack of a cloud application. The stated goal is to avoid the complexity, delays and risks of connecting to third-party services designed for production.

Waldemar Hummer, co-founder and CTO of LocalStack, sums up the intention: "Cloud applications rarely operate in isolation. With WonderTwin AI, we are extending local development beyond cloud infrastructure, to the third-party APIs and external services that applications depend on." Tela Andrews, founder of WonderTwin AI, is more blunt: "The testing environment for SaaS applications has always been limited. AI agents have completely broken it."

A Swiss startup with an American structure

LocalStack was founded by Gerta Sheganaku (COO and co-CEO), Waldemar Hummer (CTO and co-CEO) and Thomas Rausch (Head of Engineering). Gerta Sheganaku and Waldemar Hummer are based in Zurich, as are other members of the team. Legally, the Swiss company LocalStack GmbH is a subsidiary of StackLift Inc., a Delaware corporation.

This set-up is nothing unusual for a young software company targeting a global market, and it qualifies the Swiss startup label: the operational base is in Zurich, the parent company in the United States. Above all, the acquisition shows a team partly based in Zurich choosing to grow through acquisition rather than developing these emulators on its own, in a specific niche: developer tooling in the age of AI agents. It remains to be seen, beyond the promises of the press release, how customers will adopt these extended sandboxes.

Sources

  • Startupticker.ch: "Localstack strengthens capabilities with acquisition of WonderTwin AI"