Workflows
Workflows orchestrate the document intelligence pipeline: ingestion, context-aware extraction, enrichment, chunking, and delivery. They run on a schedule, on demand, or when new data arrives, so your Vector Catalog and downstream systems stay up to date without manual runs. Bundata Workflows help teams turn document processing into repeatable production systems — from legal documents and contracts to receipts and operational files.
Overview
Workflows are the orchestration layer for document AI. They provide:
- Source-aware ingestion — Trigger from storage, uploads, integrations, or external systems.
- Extraction and validation — Run schema-aware extraction and verify outputs before publication.
- Vector publishing — Send approved smart bites into the Vector Catalog and vector-ready collections.
- Operational delivery — Push outputs to agents, search, APIs, webhooks, and downstream workflows.
Scheduled jobs, smart-routing of documents, and optional auto-workflow optimization keep AI outputs in sync with the latest business content. Platform: Platform → Workflows. Product: Product → Workflows.
Key tasks
| Task | Guide |
|---|---|
| Understand what workflows do and how they fit the platform | Overview |
| Configure schedule and event-based triggers | Triggers & scheduling |
| Inspect run history, retries, and failures | Monitoring |
Tutorials and concepts
- Run your first workflow — Create a workflow with a trigger and extraction step.
- Extraction overview — What runs inside a workflow step.
- Vector Catalog overview — Common destination for workflow output.
Related product areas
- Extraction — Core step in most workflows.
- Vector Catalog — Where workflow output is often published.
- Integrations — Source and destination connectors used by workflows.
- Agents — Consume catalog data kept fresh by workflows.