Analytics

TimeBack provides analytics infrastructure through an MCP server for AI-driven exploration of curated analytics views containing student learning data, session patterns, and performance metrics.

Prerequisites

Before using analytics:

Request access via email to timeback@trilogy.com
Verify your credentials are configured for your environment (staging or production)

MCP for AI-Driven Exploration

The Model Context Protocol (MCP) server enables AI assistants to explore and query analytics data autonomously. LLMs can discover available entities, understand relationships, and execute SQL queries without human intervention.

Use MCP if:

You want AI assistants to explore data independently
You need LLMs to answer analytical questions with live data
You're building AI-powered analytics workflows

How it works: The MCP server connects to a simplified data model—curated SQL views built on top of TimeBack's internal database. These views expose clean entity relationships. Queries are limited to 500 rows and 110 seconds execution time.

For the wider choice between REST, MCP servers, and webhooks, see Choosing an Integration Surface. For per-student coaching insight reads (not aggregate analytics views), see Level 4: Insights API.

Data Architecture

The Analytics MCP queries curated analytics views:

Student data: enrollments, mastery, XP
Session data: timing, duration, waste patterns
Content data: courses, lessons, tests, questions
Performance data: attempts, correctness, engagement

These views are built on top of TimeBack's internal database and expose clean entity relationships. See Analytical Entities for complete field documentation and join keys.

Analytical Entities

Complete catalog of all analytics entities with field definitions and join keys.

MCP Tools

Programmatic API for executing queries and fetching metadata.

Analytical Entities