Schedule a Demo
Solutions How It Works API & MCP Customers About
Schedule a Demo Login
API & MCP · Developer Platform

Connect your AI agents to the learning loop.

Time Machine now exposes a full REST API and MCP server. Your AI agents and automations can provision learners, build courses, run role-plays, certify skills, and read readiness scores - so training stops being a one-time project and becomes a continuous, autonomous loop your company owns.

Schedule a Demo Read the API docs
116Agent tools (MCP)
18REST endpoint groups
REST + MCPTwo ways to connect
YoursYour IP, your data
Why this matters

Human capital + token capital, compounding.

"You can offload a task, or even a job, but you can never offload your learning. The future of the firm is the ability to compound that learning across people and AI." - Satya Nadella, Chairman & CEO, Microsoft

Every company now builds two kinds of capital: human capital - the judgment and relationships of its people - and token capital - the AI it owns. The edge isn't picking the best model; it's the learning loop on top of the models, where both compound and the expertise stays yours.

That loop is Time Machine, and the API and MCP make it programmable: your agents orchestrate; Time Machine supplies the simulations, certifications, and scorecards that turn knowledge into measurable skill. Swap the model anytime - the "company-veteran" expertise lives in the system you own.

The capabilities

What your agents can do through Time Machine.

One API surface covering the entire learning lifecycle - from provisioning a new hire to certifying them and reading their readiness. Every capability below is callable over REST or as an MCP tool.

Users & provisioning

Create, update, deactivate and reactivate team members the moment they join - or the moment HRIS says so.

Users

Departments & roles

Model your org structure and assign people to roles, so training paths map to real responsibilities.

Departments

Courses & versions

Create empty courses, assign roles, publish, and read published version snapshots - fully programmatically.

Courses · Course Versions

Modules, topics & lessons

Structure curriculum end to end - modules, topics, learning bytes, lessons and intros - and reorder it on the fly.

Modules · Topics · Lessons

Role-plays & exercises

Generate text, voice, and video simulations (ILE / RPE) and demo & app trainings where reps practice for real.

Exercises · Demo Trainings

Quizzes & knowledge checks

Author quiz questions and in-lesson checks (multiple-choice or open-ended) to verify understanding, not just completion.

Quiz Questions · Lesson Questions

AI experts

Ask company AI experts synchronous questions over the API - grounded in the documents you assign to them.

Experts

Documents & media

Upload, replace and link source material - docs, PDFs, decks, images and recordings - then attach it to topics.

Documents · Media · Topic Documents

Certificates & insights

Read certifications and PDFs, plus company- and student-level insights, scorecards and activity events.

Certificates · Insights
Model Context Protocol

One MCP server. Every model.

Point Claude, ChatGPT, Gemini, or your own agent framework at the Time Machine MCP server and the entire capability set above becomes a set of agent tools. Your agent coordinates training in plain language - "onboard this new AE," "check if Jordan is ready for live calls," "regenerate the demo course from the new release notes" - and Time Machine does the work.

  • Plain-language access to the same operations as the REST API
  • Drop-in for Claude Code, Claude Desktop, and MCP-aware agents
  • Same API-key auth and the same permissions as REST
  • Lets your existing AI agents own orchestration; Time Machine owns teaching, evaluating and certifying
Claude Claude Code ChatGPT Gemini Cursor Custom agents
Agent · plain language → MCP tools 
# Prompt to your MCP-connected agent
"A new AE, Jordan Vega, just started on the
 Enterprise team. Onboard them and tell me
 when they're certified to take live calls."

# The agent calls Time Machine MCP tools
tml_create_user(email="jordan.vega@acme.com",
                firstName="Jordan", lastName="Vega")
tml_assign_user_to_role(user="jordan.vega",
                        role="account-executive")
tml_get_course_insights(course="AE Onboarding")
tml_get_certificate(user="jordan.vega")

# → "Jordan is 80% through onboarding.
#    Objection-handling role-play: needs work.
#    Not yet certified for live calls."

Connect in under a minute

The MCP server lives at https://api.timeml.ai/api/v1/mcp (Streamable HTTP). Drop this into your client's config and authenticate with the same tml_live_ key you use for REST.

Claude Desktop · Cursor · Cline — mcp config 
{
  "mcpServers": {
    "timeml": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://api.timeml.ai/api/v1/mcp",
        "--header",
        "Authorization: Bearer tml_live_<your-key>"
      ]
    }
  }
}
Claude Code (CLI) 
# Add the Time Machine MCP server in one command
claude mcp add --transport http timeml \
  https://api.timeml.ai/api/v1/mcp \
  --header "Authorization: Bearer tml_live_<your-key>"

# Then just ask, in plain language:
#   "What tools do you have from the timeml server?"
#   "List the courses in my company."

Stdio-only clients (like Claude Desktop) reach the HTTP endpoint through the open-source mcp-remote bridge; HTTP-native clients point straight at the URL. Every call is scoped to your company and audited per key.

Quickstart

Authenticate once. Then drive the loop.

Issue a key from Settings → API Keys, send it as a bearer token, and you're live. Base URL https://api.timeml.ai, endpoints under /api/v1.

Authentication 
# Every request carries your API key as a bearer token
curl https://api.timeml.ai/api/v1/users \
  -H "Authorization: Bearer tml_live_<your_token>"
Provision a new learner 
POST /api/v1/users
curl -X POST https://api.timeml.ai/api/v1/users \
  -H "Authorization: Bearer tml_live_<token>" \
  -H "Content-Type: application/json" \
  -d '{
    "email": "jordan.vega@acme.com",
    "firstName": "Jordan",
    "lastName": "Vega",
    "departmentId": "sales",
    "roleId": "account-executive"
  }'
Read readiness for a student 
GET /api/v1/insights/students/{userId}
curl https://api.timeml.ai/api/v1/insights/\
students/jordan-vega \
  -H "Authorization: Bearer tml_live_<token>"

# → scorecards, certifications,
#   role-play performance, skill gaps,
#   readiness scoring
Ask an AI expert 
POST /api/v1/experts/{id}/ask
curl -X POST https://api.timeml.ai/api/v1/\
experts/sales-playbook/ask \
  -H "Authorization: Bearer tml_live_<token>" \
  -H "Content-Type: application/json" \
  -d '{
    "question": "How do we handle a security objection?"
  }'
Publish a regenerated course 
POST /api/v1/courses/{id}/publish
curl -X POST https://api.timeml.ai/api/v1/\
courses/product-101/publish \
  -H "Authorization: Bearer tml_live_<token>"

# Pair with documents + experts endpoints
#   to rebuild a course whenever new
#   product docs are approved.
Use case · Engineering

Headless onboarding, fully automated.

A real customer pattern: an automation orchestrated by Claude Code runs daily and wires Time Machine into the stack - new hires get provisioned and trained with zero manual steps.

The stack

  • Claude Code - orchestrator that ties the systems together
  • Time Machine - the learning experience and certification engine
  • NetSuite (HRIS) - the source of truth for new employees
  • Microsoft Dynamics + Office 365 - CRM, email, Teams, OneDrive
  • Power Automate - human-in-the-loop approvals

When HR adds an employee to the HRIS, the automation provisions accounts everywhere - including Time Machine - creates the department and role if missing, and enrolls the hire in their welcome and tech-stack courses. Weekly, it syncs learning stats back to the HRIS and nudges anyone behind.

1
New hire entered in HRISNetSuite triggers the daily automation.
2
Accounts provisioned everywhereDynamics, Office 365, and Time Machine via POST /users. Department & role created if missing.
3
Courses auto-assigned"Welcome" + "Tech Stack: Demo & App Training" enrolled on day one.
4
Progress synced & nudgedGET /insights writes weekly stats to HRIS; reminders go out over Office 365.
5
Content stays currentNew product doc in OneDrive → courses & experts regenerate and republish automatically.
Use case · Sales

The sales manager who stopped chasing spreadsheets.

This team had already built its own AI ecosystem - agents wired to CRM, playbooks and SOPs. The problem wasn't information, it was execution. Time Machine became the learning and certification engine behind it.

From "did they finish?" to "are they ready?"

Their AI agent connects over MCP and API. When a rep joins, it builds a personalized path from role, territory and experience - then monitors progress, answers questions, assigns remediation when they struggle, and advances them on mastery.

Instead of "did they finish onboarding?" the manager gets objective answers to "can they handle objections, deliver the demo, and own the conversation?" - straight from Time Machine scorecards, assessments and role-plays.

  • Personalized paths from CRM + experience data
  • Continuous coaching and automatic remediation
  • SCORM content tracked alongside AI learning
  • RACI-based accountability per learning objective
  • Readiness verified before reps touch a live deal

Automated readiness report → manager

↑
Time-to-productivityTrending down quarter over quarter.
✓
Certification progress9 of 11 new reps certified on discovery.
!
Skill gap detectedObjection handling below bar for 3 reps → remediation auto-assigned.
◎
Coaching prioritiesRanked list of who needs the manager's time, and why.
⤓
Exported to BI / CRM / HRISReadiness data flows into the systems leadership already uses.
Endpoint reference

18 groups. The whole learning lifecycle.

A map of the public REST API. Full request/response schemas live in the interactive OpenAPI docs.

Users 6

Create, read, update, deactivate and reactivate team members.

POST /api/v1/users PATCH /api/v1/users/{id}

Departments 7

Manage departments, create roles, and assign users to roles.

POST /api/v1/departments POST /api/v1/departments/{id}/roles

Courses 9

Create, update, publish and role-assign company courses.

POST /api/v1/courses POST /api/v1/courses/{id}/publish

Course Versions 2

Read published version snapshots and their content.

GET /api/v1/courses/{id}/versions

Course Modules 7

Create, edit, reorder and delete modules inside a course.

POST /api/v1/courses/{id}/modules

Course Topics 8

Create, edit, reorder, move and delete topics inside modules.

POST /api/v1/modules/{id}/topics

Learning Bytes 5

Group lessons into learning bytes inside a topic.

POST /api/v1/topics/{id}/learning-bytes

Lessons 5

Create, edit and delete lessons inside a learning byte.

POST /api/v1/learning-bytes/{id}/lessons

Intros 6

Short narrative shown before a module or topic.

PUT /api/v1/topics/{id}/intro

Exercises 7

Role-play exercises: ILE_TEXT, RPE_VOICE and RPE_VIDEO.

POST /api/v1/topics/{id}/exercises

Quiz Questions 12

Quiz questions attached to a module or topic.

POST /api/v1/topics/{id}/quiz-questions

Lesson Questions 6

In-lesson checks: MULTIPLE_CHOICE or OPEN_ENDED.

POST /api/v1/lessons/{id}/questions

Demo Trainings 7

Demo & application training videos, availability and reprocessing.

POST /api/v1/demo-trainings

Experts 8

List experts, assign documents and roles, and ask questions.

POST /api/v1/experts/{id}/ask

Documents 6

Upload, replace, update and delete company documents.

POST /api/v1/documents PUT /api/v1/documents/{id}

Media 6

Manage company images and videos.

POST /api/v1/media

Topic Documents 3

Link and unlink documents to and from a topic.

POST /api/v1/topics/{id}/documents

Certificates 3

Read earned certificates and download their PDFs.

GET /api/v1/students/{userId}/certificates

Insights 3

Company- and student-level learning insights and events.

GET /api/v1/insights/company GET /api/v1/insights/students/{userId}
Built for agents and humans

Designed to be read by machines, trusted by people.

This platform is built for a world where AI agents are first-class users - without compromising the control and ownership that make the learning loop your IP.

Two ways in: REST + MCP

A clean OpenAPI 3.1 REST API for code, and an MCP server for agents that prefer plain language. Same operations, same auth.

Structured data & llms.txt

Schema.org markup and a maintained /llms.txt so answer engines and crawlers can describe this API accurately.

Plain-language descriptions

Every endpoint group is named and described for humans first - so an agent reading the docs knows exactly what each tool does.

Your IP stays yours

Swap the underlying model anytime without losing the company-veteran expertise encoded in your learning system. Sovereignty by design.

Common questions

API & MCP, answered.

What is the Time Machine API and MCP server?

The Time Machine API is a public REST API (OpenAPI 3.1) that lets your systems and AI agents drive the learning loop: provision learners, build and publish courses, generate role-play simulations, query AI experts, certify skills, and read readiness and insight data. The MCP server exposes those same capabilities to AI assistants like Claude, ChatGPT and Gemini as agent tools, so an agent can coordinate training in plain language.

How do I authenticate?

Issue an API key from Settings → API Keys. Keys are prefixed tml_live_ and sent as an HTTP bearer token: Authorization: Bearer tml_live_<token>. The REST base URL is https://api.timeml.ai (endpoints under /api/v1); the MCP server is at https://api.timeml.ai/api/v1/mcp.

What can an AI agent actually do?

Create and manage users, departments and roles; build courses, modules, topics, lessons and versions; author quizzes and AI role-play exercises (text, voice, video); ask AI experts grounded in your documents; upload and link documents and media; and read certificates, scorecards and per-student readiness insights. There are 18 endpoint groups covering the full learning lifecycle.

Do we keep ownership of our training data and IP?

Yes. Time Machine is the learning loop you own. Your expertise, courses, role-plays, assessments and outcome data stay yours. You can swap the underlying generalist model without losing the company-veteran expertise built into your learning system - which is the real test of control and sovereignty in the AI era.

Can agents drive both onboarding and ongoing coaching?

Yes. Agents create personalized onboarding paths when a person joins, then continuously monitor progress, assign remediation when a learner struggles, advance them on mastery, and surface readiness and coaching priorities to managers - turning a one-time setup into a continuous, autonomous training loop.

Is the API ready for production use today?

Yes - the public REST API and MCP server are live. The fastest way to scope your integration is to talk to us: we'll map your agents and systems to the right endpoints and help you stand up the first workflow. Browse the interactive docs anytime.

Get API & MCP access

Put your agents to work on real training.

Tell us about your stack and what you want to automate. We'll map it to the API and MCP and come back within one business day.

REST + MCP · SOC 2 Type II · Your IP stays yours · Or call 949-416-5055