TraceOps AI: Building an AI-Powered Service Investigation Agent with FalkorDB and MCP

Description and use

TraceOps AI leverages FalkorDB, MCP, and a local Qwen model to quickly surface interconnected data about services, incidents, and runbooks, easing engineers’ investigation of microservice failures.

Problem

Engineers struggle to understand connected operational context when a microservice fails, making it hard to identify impacted services, ownership, incidents, deployments, and runbooks.

Outcome

The prototype demonstrates that a graph database combined with MCP‑exposed tools and a local LLM can efficiently surface connected operational evidence, reducing investigation time and improving clarity of root‑cause analysis.

Workflow

• Receive user question about a service
• Agent interprets request and selects relevant operational data
• MCP tools query FalkorDB for service details, dependencies, ownership, incidents, runbooks, and deployment correlations
• Graph queries traverse relationships, including multi‑hop dependencies
• Agent constructs investigation response using retrieved evidence
• Frontend displays the investigation output to the user

Evidence

The goal is not to claim that every investigation establishes a definitive root cause. Instead, TraceOps AI aims to help connect relevant evidence and distinguish observed relationships from conclusions that still need verification.

Source

The prototype includes synthetic operational data representing services, teams, incidents, deployments, and runbooks.

Source

For this project, I integrated graph investigation capabilities through the Model Context Protocol (MCP).

Source