Docs map¶
This page is the canonical reference for Docs map.
Who this is for¶
- Anyone evaluating the RecSys suite for adoption
- Teams onboarding new engineers, analysts, or on-call staff
What you will get¶
- A “pick your path” map: if you are X, read Y
- The shortest links to the tutorial, core explanations, and integration/operations references
Pick your path (by goal)¶
- Prove the loop locally (copy/paste): local end-to-end (service → logging → eval)
- Do a production-like ship/rollback run: production-like run (pipelines → object store → ship/rollback)
- Integrate the API into your product: How-to: integrate recsys-service into an application
- Troubleshoot integration issues: Troubleshooting for integrators
- Deploy on Kubernetes (Helm): Deploy with Helm (production-ish)
- Operate pipelines day-to-day: How-to: operate recsys-pipelines
- Evaluate and make ship/rollback decisions: How-to: run evaluation and make ship decisions
Pick your path (by role)¶
Developer / platform engineer¶
- Architecture + boundaries: Suite architecture
- Config + contracts:
- Configuration reference
- Data contracts
- API schema (Swagger UI): API Reference
Recommendation engineer / ML engineer¶
- RecSys engineering hub
- Ranking core (ports/adapters, signals): recsys-algo
- Pipelines lifecycle + artifacts: recsys-pipelines docs
Product / business stakeholder¶
- What you get + what you need: What the RecSys suite is (stakeholder overview)
- Pilot plan and ownership: Pilot plan (2–6 weeks) and Responsibilities (RACI): who owns what
- How success is measured: Metrics: what we measure and why
SRE / on-call¶
- Suite runbooks: Runbook: Service not ready
- Pipeline runbooks: Runbook: Pipeline failed
- Capacity guidance: Performance and capacity guide
Read next¶
- Start here (overview): Start here
- Stakeholder overview: What the RecSys suite is (stakeholder overview)