Let your AI assistant request real-world services

From intent to a real service request in one tool call

Expose the API as a tool or function in your agent framework. The model collects category, location and description, calls the endpoint, and confirms to the user that the request has been submitted to the marketplace.

• Works as a function/tool call in any agent framework
• Structured payload, predictable responses
• Audio requests supported for voice assistants
• Duplicate validation avoids repeated requests from chatty agents

Real outcomes, not recommendations

Recommending "call a plumber" is where most assistants stop. Creating the request and reaching professionals who are actually active in that category and location is what users want - and what turns your assistant into a product people pay for.

Reviewed and private by design

Access is private and approved per partner, requests are rate-limited and validated, and every integration is reviewed before going live to protect the experience of end users and professionals alike.

Frequently asked questions

Can an autonomous agent call the API without human confirmation?

Technically yes, but we recommend a user confirmation step before creating a request, and we review this flow during the integration process.

Does the API accept voice input?

Yes. Audio service requests (mp3/m4a, base64) are available to approved partners for voice assistants.

Can we prepare our integration before going live?

Yes. Technical documentation and integration resources are provided to approved partners so your team can build and test the integration before launch.