Tools + ontology
Your job, drawn as a graph.
Run it as a checkpointed DAG. Inspect every node.
The problem
A long workflow is a black box.
If step six goes wrong, you find out at step ten.
You can't price what you can't see.
What it does
Plan it. Price it. Check it.
- See the plan before the job runs.
- Price it. A job you can plan is a job you can quote.
- Inspect every node's output.
- Resume, don't restart. Each step is checkpointed.
- Your data is stored in the ontology manifold.
A job drawn as a graph
Each step's result is savedEvery node can be inspected
Illustrative: a job drawn as a graph.
Questions
Quick answers
What is a checkpointed DAG?
A graph of steps where each step's result is saved, so a job can be inspected and resumed instead of restarted.
Why does a graph help with pricing?
A job drawn as a graph has steps you can see and plan before it runs.
Sound familiar?
Pipelines nobody can see into.
The system