> For the complete documentation index, see [llms.txt](https://intelfactor.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://intelfactor.gitbook.io/docs/intelfactor-ai/architecture-and-data/edge-pipeline.md).

# Edge pipeline

Conceptual responsibilities and execution guarantees at the edge.

Defines what the edge pipeline does and what it guarantees.

### Responsibilities of the Edge Pipeline

The edge pipeline is responsible for:

```mermaid
sequenceDiagram
  autonumber
  participant Cam as Camera
  participant Edge as Edge service
  participant Infer as Inference
  participant Policy as Policy/guardrails
  participant Store as Local store (DB+FS)
  participant UI as Local UI/dashboard
  participant Site as Site hub (optional)

  Cam->>Edge: Frame/stream
  Edge->>Infer: Preprocess + infer
  Infer-->>Edge: Detections + confidences
  Edge->>Policy: Apply thresholds + rules
  Policy-->>Edge: PASS / FAIL / REVIEW
  Edge->>Store: Write event metadata + evidence refs
  Edge->>UI: Update station status + counts
  Edge-->>Site: Sync events/evidence (optional, async)
```

* Ingesting live video or sensor streams
* Evaluating inspection conditions in real time
* Producing inspection outcomes
* Capturing inspection evidence
* Emitting structured inspection events

### Design Characteristics

The edge pipeline is designed to be:

* Deterministic in latency
* Bounded in resource usage
* Configurable per inspection station
* Predictable under load

Specific algorithms, models, and optimization strategies are proprietary and intentionally omitted.

### Next steps

* [Local RCA engine](/docs/intelfactor-ai/product-concepts/local-rca-engine.md)
* [Safety rails](/docs/intelfactor-ai/product-concepts/safety-rails.md)
* [Interfaces](/docs/intelfactor-ai/architecture-and-data/interfaces.md)