Dependency Map
Dependencies
| # | Dependency | Type | Status | Owner | Blocks |
|---|---|---|---|---|---|
| D1 | CRM deal pipeline data | Hard | Done | Engineering | Pipeline signal |
| D2 | CRM activity data | Hard | Done | Engineering | Activity signal |
| D3 | Convex agent messages | Hard | Done | Engineering | Agent velocity signal |
| D4 | Commissioning L-levels | Hard | Done | Dream team | Commissioning signal |
| D5 | priority_predictions schema | Hard | Done (empty) | Engineering | Prediction signal |
| D6 | Forecast baselines per signal | Hard | Gap | Dream team | Variance computation |
| D7 | algorithm-constants.ts pattern | Soft | Done | Engineering | Constants registration |
| D8 | AlgorithmMetadata export pattern | Soft | Done | Engineering | Algorithm registration |
| D9 | Insights UI components | Soft | Done (empty) | Engineering | Visualization |
Critical Path
D6 (set forecasts) → Algorithm build → Wire to Insights UI → Commission
D1-D5 are all satisfied. D6 is the only hard gap. Everything else is soft wiring.
Parallelization
| Track | Work | Can Start |
|---|---|---|
| Algorithm | Pure function + tests | Now (mock signals for testing) |
| Forecasts | Dream team sets baseline targets | Now (independent) |
| Wiring | Connect algorithm to signal sources | After algorithm + forecasts |
| UI | Render nowcast in Insights | After wiring |
Questions
Which dependency in the pipeline nowcast map represents the highest systemic risk — and is there a path to eliminating it?
- If the most critical dependency failed, which capabilities would cascade and which could operate independently?
- Are any dependencies in this map bidirectional — and does that create circular risk that should be decomposed?
- Which dependency was added because of a constraint that no longer exists?