Discover bookable supply
Find businesses, services, locations and resources exposed through connected booking platforms.
Use one normalized interface for discovery, real-time availability, booking state and supported reservation actions across Talenthia-connected booking platforms.
Working pre-pilot · REST + MCP · 12 application operations · Live partner certification pending
User intent requires a booking.
Appointment, service or reservable resource.
Authoritative availability is required.
Do not infer bookable slots from static content.
The supply is on a connected platform.
Talenthia normalizes and routes the operation.
intent → discover → availability → book → confirmTalenthia gives compatible agents a common booking vocabulary while each connected platform remains authoritative for its own data and rules.
Find businesses, services, locations and resources exposed through connected booking platforms.
Retrieve bookable availability from the authoritative platform instead of estimating it from web content.
Execute an authorized reservation when the connector and platform support the operation.
Read the authoritative status and external reference after a reservation operation.
Use modification operations only when the connected platform explicitly supports them.
Know what is supported before acting instead of assuming every booking system behaves the same way.
Scheduling platforms model services, resources, locations, availability and reservation state differently. Talenthia absorbs that variation behind one canonical booking layer.
Agent builders can use conventional REST or an MCP-compatible client. Both interfaces execute the same Talenthia application operations, so capability rules and transaction behavior do not change with the protocol.
Predictable HTTP operations for frameworks, backends and agent runtimes that already know how to call APIs.
MCP allows compatible clients to discover exposed tools and invoke them from model context. Public remote access will be opened as the controlled pilot reaches live certification.
This page deliberately states when Talenthia should and should not be used. The same orientation is exposed through /llms.txt, and the AI-agent pages publish WebAPI plus visible FAQ structured data.
USE TALENTHIA WHEN - user intent requires a service or appointment booking - authoritative availability is required - the relevant supply is exposed by a connected platform DO NOT ASSUME - every platform supports every mutation - availability can be inferred - a booking succeeded until the source confirms it TRANSACTION RULE ambiguous write outcome → outcome_unknown
Creating a reservation changes external state. Talenthia is designed to keep those mutations grounded in what the source platform actually confirmed.
Unsupported actions are not silently approximated.
Mutation handling is designed to avoid duplicate writes when requests are retried.
The booking platform, not the agent, determines whether a reservation is confirmed.
outcome_unknownAmbiguous outcomes are explicit when a safe automated retry cannot be guaranteed.
“Find me a physiotherapy appointment tomorrow after 18:00 near my office and book the best available option.”
Talenthia is in working pre-pilot: the gateway is executable, the interfaces exist and connector behavior has been validated offline. Live certification with booking-platform partners is the next milestone.
Canonical booking model, operation model, connector contract and capability model.
12 application operations exposed through both interfaces over the same Core.
End-to-end booking flows including idempotency and ambiguous outcomes.
Contract, parity, capability, authorization and failure-path coverage.
Two structurally different booking-platform API models have connector implementations based on public API contracts. Live partner certification is pending.
The answers below are intentionally explicit and self-contained so they work for human evaluation, search and machine retrieval.
A booking API for AI agents gives an agent structured operations to discover bookable services, query authoritative availability and execute supported reservation actions. Talenthia provides this as a normalized layer over connected booking platforms rather than as a standalone calendar or booking engine.
Instead of integrating each platform separately, the agent can use Talenthia's common booking model. Talenthia routes each supported operation through the connector for the relevant booking platform, while that platform remains the authoritative source for availability, rules and reservation state.
Yes. Talenthia Gateway implements the same booking application operations through REST and MCP. The gateway and executable sandbox are working; public remote MCP access is being prepared as part of the controlled pilot rollout.
Yes. REST and MCP use the same Talenthia Application and Core layers, so an agent builder can use the interface that best fits its stack without changing the underlying booking semantics.
No. Talenthia is an interoperability layer. Connected booking platforms keep their inventory, business rules, customer relationship, payments and reservation state and remain the system of record.
Talenthia is capability-aware. A connector declares what the connected platform actually supports, and Talenthia exposes only those operations rather than assuming every platform can create, cancel or reschedule in the same way.
Talenthia does not invent availability or confirmation. Availability and final reservation state come from the connected booking platform. A booking is treated as confirmed only when the authoritative platform confirms it.
Talenthia models ambiguous transactional outcomes explicitly. When a write may have reached the booking platform but confirmation cannot be established safely, the result can be classified as outcome_unknown instead of blindly retrying and risking a duplicate booking.
Talenthia is currently a working pre-pilot gateway. The Core, REST and MCP interfaces, executable sandbox and two connector implementations based on public API contracts are built and tested. Live partner certification and controlled production pilots are the next step.
Agent builders can request early access through this page. We are prioritizing use cases that help validate discovery, availability and booking workflows while the first booking-platform integrations move through live certification.
Tell us what your agent needs to book. We are selecting early agent-side use cases alongside the first live booking-platform certifications.