Skip to content

Add align pipeline node #41

@natashaannn

Description

@natashaannn

Add align pipeline node

Objective

Migrate transcript alignment into the DAG.


Why this matters

Alignment is one of the most important deterministic transcript stages.

Adding it makes the pipeline meaningfully closer to production.


Parallelization metadata

Track

pipeline

Depends on

Safe to run in parallel with

  • diarize node
  • shared typing issues
  • remotion utility issues

Merge risk

  • medium

Files in scope

Primary:

  • scripts/pipeline/nodes/align.ts

Avoid touching:

  • WhisperX implementation details

Required implementation

1. Implement node wrapper

Expose align as a PipelineNode.

2. Define inputs/outputs

Artifact references only.

3. Delegate to existing alignment behavior

No alignment algorithm changes.


Constraints

  • wrapper only

Handoff contract

Future stages can depend on aligned transcript artifacts.


Acceptance criteria

Functional

  • align executes through runner

Integration checks

  • aligned transcript artifact is created

Verification commands

Run pipeline through align.

Expected result:

  • aligned transcript exists

Explicitly out of scope

  • assign-speakers
  • merge-doc

Suggested branch

refactor/s6-align-node

Suggested commit slug

phase-2-step-2-align-node

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions