# Thomas Developers — Overview

The validated people-data layer, for partners, platforms and developers. Embed 40+ years
of validated psychometrics and the Thomas Connection model into your product so your
customers can understand and support their people across the whole employee journey:
hiring, onboarding, communication, development, role changes and departures.

Any product, any language. A first integration ships in an afternoon.

## The golden path: resolve → assign → read

One flow, from people to your product. Your people take Thomas assessments; Thomas turns
them into validated intelligence; you consume it wherever you need it.

1. **Connect your people** — resolve or import your directory. Each person is matched by
   email or HRIS id. Operation: `resolveIdentity`
   ([POST /v1/individuals/resolve](/docs/api/operations/resolve-identity.md)).
2. **Trigger an assessment** — invite a candidate or employee, or surface it inside your
   own flow. Operation: `createAssessmentAssignment`
   ([POST /v1/assessments/assignments](/docs/api/operations/create-assessment-assignment.md)).
3. **Use the intelligence** — pull profiles, connection and coaching into your product,
   for one person or a whole team. Operation: `getAssessmentBehaviourProfile`
   ([GET /v1/assessments/profile/{individualId}](/docs/api/operations/get-assessment-behaviour-profile.md)).

## What partners build

- **Hire smarter** — rank candidates against a role on validated behaviour, aptitude and EQ, inside your ATS.
- **Guide managers** — surface how to work with and communicate with each report.
- **Help people progress** — recommend the next learning step from a person's profile, in your LMS.
- **Run better reviews** — enrich performance reviews with connection and behaviour insight.

Not only for HR: any product with people in it (a CRM, a collaboration tool, a community)
can embed Thomas. People complete an assessment once, then the profile is available
everywhere you build.

## Built for volume

Bulk endpoints, validated up front and idempotent by email. A full directory of 10,000
people is about 100 batched calls, not 10,000. Batches of up to 100, validation up
front, EU infrastructure.

## For AI agents

The same OpenAPI artifact that renders these docs is what an AI agent reads to answer
"can we integrate Thomas, and how?" accurately, on a developer's behalf.

- Agent map: [/llms.txt](/llms.txt)
- Full agent context: [/llms-full.txt](/llms-full.txt)
- OpenAPI artifact: [/openapi/thomas-federated.openapi.json](/openapi/thomas-federated.openapi.json)

## Watch it work

Before you read a single doc, run the inline Discovery demo right here on the Overview
page: pick a seeded question and watch a Thomas-powered agent work through it, step by
step. It runs against seeded people and real Thomas response shapes, with no signup and
no network call to a real tenant.

The same seeded demo is also a direct, shareable route at [/discovery](/discovery),
for deep links, prerendering and crawlers.

## Where next

- [Quickstart use cases](/quickstart.md)
- [API reference](/api-reference.md)
- [Concepts](/concepts.md)
- [Trust & compliance](/trust.md)
- [Discovery demo](/discovery)
