Skip to main content

​
What is Brixo?

Brixo helps Product, Support, Sales, and Executive teams understand what’s really happening inside AI-powered conversations. Brixo answers three fundamental questions:
  1. Why users engage with AI agents — their goals and intent
  2. How users experience those interactions — sentiment, effort, friction
  3. What outcomes those interactions produce — resolution, containment, deflection, CSAT
Brixo bridges the gap between engineering telemetry and business outcomes by transforming raw conversational data into structured signals and KPIs teams can act on.
Brixo is not an engineering observability tool.
We use telemetry as an input to measure human experience and business impact.

​
What Brixo Measures

Brixo analyzes conversations using two layers:

​
Signals (what happened)

  • User Goals (e.g. Reset Password, Cancel Subscription)
  • Conversation Topics (e.g. Billing, Technical Support)
  • User Sentiment
  • Conversational Events (loops, escalations, abandonments, wins)

​
Metrics (why it matters)

  • Resolution Rate
  • First Contact Resolution (FCR)
  • Effort Score
  • CSAT / Health
  • Containment / Automation Success
  • Deflection Rate
  • Average Handle Time (AHT)
These metrics map directly to executive dashboards, operational reporting, and ROI analysis.

​
Choose Your Integration Path

Brixo supports two first-class ways to send conversation data.
Both are valid — choose based on how your system is built today.

​
Path 1 — Use the Brixo SDK (Explicit Conversation Instrumentation)

Best for teams who want maximum accuracy and control over how conversations are represented. With the Brixo SDK, you explicitly define:
  • Interaction boundaries (one user request → one response)
  • User, account, and session identity
  • User-visible input and final agent output
This guarantees that Brixo sees the conversation exactly as the user experienced it. Recommended if:
  • You want precise resolution, effort, and health metrics from day one
  • You control the agent runtime
  • You want minimal inference and maximum correctness
→ Get started: /quickstart/python

​
Path 2 — Use Your Existing OpenTelemetry (OTLP) Integration

Best for teams who already emit OpenTelemetry traces and want to get started quickly. Brixo can ingest traces directly via OTLP (HTTP or gRPC).
You can stream your existing traces to Brixo without changing your observability stack. This path supports progressive enrichment:
  • Start by sending what you already emit
  • Layer in lightweight context over time to unlock deeper analytics
Recommended if:
  • You already use OpenTelemetry (e.g. Pydantic AI, framework auto-instrumentation)
  • You want minimal upfront integration work
  • You plan to evolve instrumentation incrementally
→ Get started: /quickstart/otlp
OTLP is the transport.
Conversation Analytics depends on whether traces reflect complete user interactions.

​
Progressive Enrichment (Applies to Both Paths)

You don’t need perfect instrumentation on day one. A common progression:
  1. Send existing data (SDK or OTLP)
  2. Validate interaction boundaries
  3. Add user/session context
  4. Add explicit user-visible input/output
  5. Unlock full experience and outcome analytics
Brixo will guide you on the smallest possible additions needed to unlock more value.

​
Which Path Should I Choose?

If you want…Choose…
Fastest start with existing telemetryOTLP Integration
Guaranteed accuracy for conversation metricsBrixo SDK
No changes to your observability stackOTLP Integration
Explicit control over conversation boundariesBrixo SDK
You can also combine both — many teams start with OTLP and later add lightweight SDK instrumentation for richer context.

​
Next Steps

  • Using OpenTelemetry? → /quickstart/otlp
  • Instrumenting an agent directly? → /quickstart/python
  • Questions? → [email protected]

​
Key Takeaway

Brixo measures conversations — not just traces.
Choose the integration path that fits your system today, and evolve from there.