# Flow: Data scientist

## Goal
Deliver a data product or model with reproducible experiments and measurable impact.

## Discovery prompts
- What business or research question are we answering?
- What data sources exist and what is their quality?
- What are the target metrics (accuracy, recall, ROI)?
- What constraints exist (latency, cost, fairness)?
- How will the model be monitored post-release?

## Required artifacts
- `requirement.md` with objective and success metrics
- `functional-spec.md` for data flows and user interactions
- `technical-spec.md` for pipelines, features, and infra
- `architecture.md` for data storage and serving
- `test-plan.md` for data validation and model evaluation

## Risk and compliance
- Data bias and fairness
- Privacy and consent
- Model drift and monitoring gaps

## Acceptance criteria examples
- Model meets minimum metric thresholds.
- Data validation passes for all critical datasets.
- Monitoring alerts exist for drift and anomalies.

## Recommended outputs
- Feature catalog
- Experiment tracking plan
- Model card summary