AI Workloads Platform¶
Carbon emissions tracking and offsetting for AI inference workloads.
Reading Order¶
Start with the Proposal for context, then follow the path that matches your goal:
| Goal | Reading path |
|---|---|
| Understanding the product | Proposal → User Stories → Requirements |
| Building it | Design → Data Model → API Contract → Tasks → Quickstart |
| Understanding decisions | Research → Design |
| Writing tests | Domain Specs → API Contract → Quickstart |
Document Map¶
| Document | Role | Summary |
|---|---|---|
| Proposal | Why and what changes | Start here for context |
| User Stories | Who uses it and how | User narratives, acceptance scenarios, success criteria |
| Requirements | What the system must do | Sections 1-11 covering all functional requirements |
| Domain Specs | How to verify it works | BDD scenarios (GIVEN / WHEN / THEN) |
| Research | Why we chose this approach | Decision log (R1-R9) with alternatives considered |
| Design | How it's built | Architecture, components, error mapping, project structure |
| Data Model | What data exists | 11 entity schemas with columns, indexes, constraints |
| API Contract | How to call it | Every endpoint with request/response shapes |
| Tasks | When to build what | Phase 1 and 2 task breakdown with requirement traceability |
| Quickstart | How to run it | Local setup, smoke tests, verification points |
Diagrams¶
Four standalone Mermaid diagrams are embedded throughout the documentation:
| Diagram | Location | Contents |
|---|---|---|
| System Context | Design | High-level system context (external services, AWS, frontend) |
| Telemetry Pipeline | Design | Data flow: polling → ingestion → emissions calculation |
| Billing & Receipt Pipeline | Design | Data flow: Stripe webhook → billing close → receipt signing → PDF |
| Entity Relationship | Data Model | Relationships across all 11 tables |
Single Source of Truth¶
Each piece of information lives in exactly one document. Other documents cross-reference rather than duplicate.
| Information | Canonical location |
|---|---|
| Functional requirements (sections 1-11) | Requirements |
| Entity schemas, indexes, constraints | Data Model |
| Architecture decisions with rationale (R1-R9) | Research |
| Implementation-level decisions | Design |
| Error-to-HTTP mapping | Design |
| API endpoint shapes and examples | API Contract |