Collect & normalize
Feed records align into a shared shape; duplicates collapse.
n8n Job Priority Tracker
AUTOMATION / DATA PIPELINESScheduled workflow · report artifacts
A useful result, even when nobody is watching.
THE CONTEXT
Make inconsistent job feeds usable for repeatable analysis.
THE KEY DECISION
Normalize and deduplicate before aggregating dated samples.
THE RESULT
A configured n8n workflow with JSON, Markdown and PDF report history.
Job Priority TrackerINPUT → REPORT
MECHANISM STUDY / 09SCHEMATIC
TAKE A CLOSER LOOKSELECT A STAGE ↓
SCROLL TO TRACE. SELECT TO EXPLORE.
Follow the connections as you scroll, or select a stage to see what happens there.
Extract & analyze
Bounded skill matching and compensation processing produce sample-level summaries.
Publish the artifacts
A scheduled runner writes JSON, Markdown and PDF reports into history.
A scheduled market-data pipeline that runs headless n8n and Node/Python processing through GitHub Actions, publishing inspectable output into the repository.
PROJECT TYPE
Workflow & analytics engineering
FOCUS
Automation + Data / analytics + Backend
STATUS
Scheduled workflow · report artifacts
/01
The question.
Different feeds represent the same job differently. Useful aggregation needs normalization, deduplication, explicit salary handling and a record of what each run produced.
/02
The approach.
Source lanes feed an n8n workflow. Records are normalized and deduplicated, bounded token rules identify skills, and Python aggregation computes sample-level trends and compensation summaries. Report generation writes JSON, Markdown and PDF artifacts with historical outputs.
/03
What came out of it.
The repository holds a configured scheduled workflow and generated report history. Report contents are dated samples; they are not a representative labor-market survey or evidence that the scheduler has run continuously.
01
Automation without a product shell
The primary interface is the recurring workflow and its outputs, rather than a hosted application dashboard.
02
Normalize before aggregating
Deduplication and consistent fields precede skill and salary analysis, making the input assumptions easier to inspect.
03
Leave a reproducible trail
Generated artifacts and dated archives expose what a run actually processed. Small samples remain small samples.
Under the surface.
- Orchestration
- GitHub Actions schedule · headless n8n
- Processing
- Node.js · bounded token extraction · Python aggregation
- Delivery
- JSON · Markdown · PDF · archived outputs
SCOPE & LIMITS
- Configured scheduling does not imply uninterrupted service.
- Small collected samples are not representative market statistics.
- The animation shows workflow stages, not live external job traffic.
NEXT PROJECT / 10
Have an idea of your own? Let’s talk