Architecture
How t2e turns telemetry into evidence — from collector to dashboard.
System Overview
t2e is composed of three main components that work together to continuously ingest, process, and present compliance evidence from your infrastructure telemetry.
Collector
A lightweight Go daemon deployed via Helm into your Kubernetes cluster. It connects to multiple telemetry sources and forwards normalized events to the backend.
POST /v1/ingest/batch Backend API
A multi-tenant FastAPI application that ingests events, runs risk rules, maps evidence to controls, generates reports, and manages organizations.
Risk Engine
The risk engine evaluates incoming telemetry against YAML-defined rules. Rules specify conditions, severity levels, and control mappings.
# rules/kubernetes.yaml
rules:
- id: K8S-001
name: "Privileged container running"
severity: HIGH
condition: payload.securityContext.privileged == true
controls:
- soc2:CC6.1
- iso27001:A.8.9
- pci_dss:2.2.1 Rules are evaluated during event normalization. Matches create risk records and link evidence to the specified controls.
Evidence Chain
Every piece of evidence is cryptographically verified:
Event received, SHA-256 hash computed
Source-specific normalizer extracts evidence payload
Evidence linked to compliance controls via YAML mappings
Append-only evidence vault with hash chain verification