ProcVue

Business Process Observability from OpenTelemetry Traces

Your infrastructure is monitored. Your business processes aren't.

ProcVue automatically builds a living map of how requests flow through your distributed system — from your existing OpenTelemetry traces. See every path, every branch, every bottleneck. In business terms, not span IDs.

🌐 procvue.com · 📬 Join Waitlist

---

The Problem

You have Grafana for metrics, Sentry for errors, Jaeger for traces. But none of them answer:

• "What percentage of payments go through Provider A vs Provider B?"
• "Where exactly do orders get stuck in the fulfillment pipeline?"
• "Did the new deployment change the ratio of retry paths?"

Teams end up building internal Grafana dashboards and SQL queries to answer these questions. It works until it doesn't — nobody maintains it, nobody trusts it, and it never shows the full picture.

How ProcVue Works

ProcVue sits alongside your existing observability stack. It reads trace data and automatically discovers business process topology — no orchestrator required, no architecture changes.

┌──────────────┐     ┌──────────────────┐     ┌────────────────┐     ┌──────────────┐
│  Your Apps   │────▶│  OTel Collector  │────▶│  ProcVue       │────▶│  ClickHouse  │
│  (instrumented)    │  (add exporter)  │     │  Ingest        │     │              │
└──────────────┘     └───────┬──────────┘     └────────────────┘     └──────┬───────┘
                             │                                              │
                             ▼                                              ▼
                     ┌──────────────┐                              ┌──────────────┐
                     │  Jaeger /    │                              │  ProcVue UI  │
                     │  SigNoz /    │                              │  (React)     │
                     │  Datadog     │                              └──────────────┘
                     └──────────────┘
                     (stays untouched)

Two Integration Paths

Path 1: Auto-discovery from OTel traces (zero-code)

Add ProcVue as another exporter in your OTel Collector config. Your existing setup stays untouched.

exporters:
  procvue:
    endpoint: https://ingest.procvue.com
    project_key: "your-project-key"

pipelines:
  traces:
    receivers: [otlp]
    exporters: [procvue, jaeger]  # keep your existing exporters

Path 2: SDK for explicit business events

For teams that want precise control over what gets tracked.

import { ProcVue } from '@procvue/sdk';

ProcVue.transition({
  entity: 'payment',
  id: 'pay_abc123',
  state: 'authorized',
  from: 'pending',
  event: 'PROVIDER_APPROVED',
  meta: { provider: 'stripe', amount: 4200 },
});

Key Features

Flow Topology — Auto-discovered process map from traces. Every path, branch point, and convergence. Traffic distribution percentages. Color-coded health.

Business Block Mapping — Group low-level spans into business blocks via DSL. Engineers see details, stakeholders see the process. Same data, different zoom level.

Live Pipeline — Real-time view: how many entities are in each state right now? Where is the queue growing? Which stage has abnormal dwell time?

Entity Inspector — Click any entity and see its complete journey. Every state transition, timestamp, and context. Jump to the raw trace in Jaeger/SigNoz for deep debugging.

Anomaly Detection — Continuous monitoring of traffic distribution across paths. Alerts when ratios shift, error rates spike, or latency creeps up on a specific branch.

Roadmap

Phase	Focus	Status
1	OTel Collector exporter → ClickHouse → Flow Topology UI	🔨 In progress
2	Business mapping DSL, Entity Inspector	⏳ Planned
3	Live Pipeline view, basic alerting	⏳ Planned
4	Anomaly detection, Slack/PagerDuty integration	⏳ Planned
5	Kafka/NATS connector, Temporal interceptor, XState middleware	⏳ Planned

Tech Stack

• Ingest: Go (OTel Collector exporter plugin)
• Storage: ClickHouse
• UI: React + TypeScript
• SDK: TypeScript (Python, Go planned)

Use Cases

• Payments & Fintech — auth → capture → settle → reconcile. Decline spikes, provider latency, retry recovery.
• E-commerce — Order lifecycle from cart to delivery. Fulfillment bottlenecks, stuck orders.
• Insurance — Claims processing with SLA monitoring per stage. Compliance audit trail.
• SaaS — Signup flows, trial-to-paid conversion, feature activation funnels (server-side).

Contributing

We're in early development. The best ways to get involved right now:

• ⭐ Star this repo to follow progress
• 📬 Join the waitlist for early access
• 💬 Open an issue with your use case or feedback
• 📧 support@procvue.com — we'd love to hear how you monitor business processes today

License

Apache License 2.0