Skip to main content

What is the Workflow API?

The Workflow API is the execution plane that orchestrates everything Beltic does — document verification, identity checks, sanctions screening, business lookups, and credential issuance — into a single callable process. Instead of manually sequencing eight API calls (create account → create session → upload document → run verification → check screening → issue credential), you configure a workflow once and call POST /v1/workflows/execute. The workflow runs the blocks in order and returns the output, including any signed credentials, in a single response.

How Workflows Are Structured

A workflow is a state graph of blocks — discrete units of work that run in sequence or in parallel. Each block type wraps a specific Beltic capability: The credential block is what turns a verification outcome into a portable, cryptographic artifact. The signed credential the workflow emits can be stored, forwarded to partners, and verified at every future transaction point — without re-running the workflow.

Synchronous vs Paused Executions

Workflow execution is synchronous by default. The HTTP response tells you the outcome:

The 202 Pause Flow

Some verification steps can’t be resolved automatically — a policy mismatch may require a manual review from a compliance officer or a sponsor bank before a credential can be issued. When this happens, the workflow pauses and returns 202 with a pausePoints array describing what input is needed to continue.
Once the external review completes, resume the execution by re-running with the runFromBlock parameter pointing to the block after the pause.

Execution Output

Every completed execution returns an output object keyed by block ID:
The signed_payload in the credential block output is the JWT-VC. Store it — it’s the artifact your downstream services verify against.

Configuring a Workflow

Workflows are configured in the Beltic Console using a visual block editor. You add blocks, connect them in order, configure each block’s inputs and conditions, and deploy — similar to how you’d configure a CI/CD pipeline, but for verification logic. What happens at deploy time: Beltic validates the workflow graph, assigns a stable workflowId (e.g. wf_kyc_standard), and creates an immutable deployed version. The deployed version is what POST /v1/workflows/execute runs by default. Earlier versions are preserved and can still be targeted via the executions API for debugging. Draft vs deployed: While editing, the workflow is in draft state. You can test draft workflows by passing "useDraftState": true to POST /v1/workflows/execute — useful for staging validation before promoting a change to production. Code-defined workflows (declarative YAML/JSON config and Git-based deployment) are on the roadmap. The REST API for workflow execution is already stable and versioned — switching from Console-deployed to code-defined workflows will not require changes to your integration.

Workflows vs Direct API Calls

The Workflow API supersedes the Identity API for most use cases. The direct comparison: The Identity API remains available for existing integrations but is not recommended for new builds.

Where to Next

Executing a Workflow

Walk through POST /v1/workflows/execute — inputs, outputs, pause handling, and partial re-runs

Credentials API

Understand the signed credential that workflows emit — and how to verify it at transaction time

Fintech Company Using Beltic

See workflows and credentials working together — from user onboarding through transaction authorization