Lovable announced on September 18, 2026 that it has acquired Sutro, the company behind the SLang programming language and a platform for generating production-ready backends. The acquisition brings Sutro’s backend technology and team into Lovable, with Sutro founder Tomas Halgas joining Lovable to lead technical evangelism.
Halgas’s role continues Sutro’s work of making the technology actually understandable to the people building with it. Sutro’s team includes Halgas and engineers Max Gfeller, Tony Zhan, and Hirad Arshadi, who bring more than 20 years of combined experience working on code-generation engines.
What Sutro Built Around SLang
Sutro markets its product as proper backends built on a zero-infra approach centered on SLang. The language defines and deploys production-grade backends, including entities, logic, and security, without boilerplate or complex syntax.
That focus gives SLang a clear job: describe the parts a backend needs, then support their deployment without requiring the usual layers of setup. Sutro’s documentation organizes the language around six core constructs: entities, schemas, actions, triggers, security, and modules.
Those constructs cover the main pieces named in Sutro’s documentation. Entities and schemas describe the backend’s structure, while actions and triggers define what it does. Security and modules complete the set, giving SLang a way to describe backend behavior and organization alongside its data and access rules.
Sutro’s platform supports the generation, hosting, and editing of production-ready backends. That combination matters because the product is not limited to creating backend code; it also covers the steps needed to host and change the resulting backend.
Two Ways to Construct a Backend
A backend can be constructed from a single prompt processed by Sutro’s AI agent. It can also be assembled manually through calls to the API, giving builders two paths into the same platform.
The prompt-based approach places the AI agent at the front of the process, while the API offers a manual route for assembling the backend. Both paths connect to Sutro’s focus on production-ready systems rather than only early prototypes or isolated code examples.
The API server provides a downloadable OpenAPI document, which describes the available API surface for a backend. The server is reachable at an address that follows this form: {applicationId}.app.withsutro.com.
That address format is part of Sutro’s platform structure, where each application uses an application ID in its server address. The downloadable OpenAPI document gives builders a direct way to work with the API definition associated with the backend.
Halgas Takes the Technical Message Forward
Halgas will lead Lovable’s technical evangelism after joining the company. The role matches the work he has already been doing at Sutro: explaining how the technology works to the people building with it.
That explanation becomes important when a platform combines a programming language, an AI agent, an API, hosting, and editing tools. Sutro’s product brings those pieces together around SLang, while its six core constructs give the language a defined structure.
The acquisition also keeps attention on the people behind the product. Sutro’s team includes Halgas, Gfeller, Zhan, and Arshadi, and the team has more than 20 years of combined experience on code-generation engines. That experience sits alongside SLang’s zero-infra approach and its support for production-grade backend features.
Lovable’s announcement therefore centers on two connected parts of Sutro: the team explaining the technology and the platform built around SLang. Sutro provides a way to generate, host, and edit production-ready backends, whether a builder starts with one prompt for its AI agent or assembles the system through API calls.
For the technology itself, the key pieces remain clear: SLang defines entities, schemas, actions, triggers, security, and modules; the platform supports backend generation, hosting, and editing; and the API server provides a downloadable OpenAPI document at an address built around an application ID.
