Databuff
GitHub
Skip to content

中文  |  English

User Guide · AI Platform

What It Is

The platform's intelligent brain — use conversation instead of digging through charts, and AI instead of manual troubleshooting.

You ask; it queries data, analyzes, and delivers conclusions.


Quick Start

  1. Configure a model: Configuration → Model Settings → enter your API key
  2. Start a conversation: AI Platform → type your question
  3. Get answers: AI automatically queries data and replies

Three Digital Experts

ExpertStrengthExample
AI BrainUnderstands your question and routes to the right expert"How is the payment service doing lately?"
Query ExpertQueries metrics, traces, topology, and alerts"QPS trend for order service"
Inspection ExpertProactively finds service anomalies"Inspect the inventory service"

In daily use, talk directly to the AI Brain — it routes automatically.


What It Can Do

ScenarioExample questions
Service health"Which services have the highest error rates?"
Performance trends"How was order service latency over the last hour?"
Slow requests"Find the 5 slowest traces"
Dependencies"Who does payment service call, and who calls it?"
Proactive inspection"Inspect all core services for me"
Incident triage"Order error rate spiked — what caused it?"

Multi-agent Collaboration

For complex questions, you don't need to break them down — the AI Brain dispatches multiple experts in parallel and returns one consolidated answer.

This is what sets DataBuff apart from a simple chatbot: it doesn't just chat — it gets work done.


Extend External Capabilities

Need to expand the platform boundary? See Custom Digital Experts — register MCP, create SPECIALIST experts, and let the AI Brain route tasks automatically.

Need GitHub, browsers, knowledge bases, or other external systems in your AI troubleshooting flow? See External MCP Integration — register remote MCP in Tool Management, bind it to a digital expert, and call it from AI chat.