Phase 1 project prototype
This commit is contained in:
parent
29215e2bd2
commit
2c9ae1c312
29 changed files with 2967 additions and 22 deletions
|
|
@ -117,10 +117,25 @@ Sprint 3: BLS CPI-U (independent) │
|
|||
+ CLI + Validation ─────────────┤
|
||||
▼
|
||||
Sprint 4: Stretch scraper ────── Phase 1 Complete
|
||||
│
|
||||
Phase 2: SKIPPED (folded into dashboard queries)
|
||||
│
|
||||
Sprint 5: Dash dashboard ──────── Phase 3 Prototype
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## Phase 2 Skip Decision (March 2026)
|
||||
|
||||
Phase 2 (data pipeline & normalization) was **skipped** for the initial prototype. All derived metrics — admin cost ratios, CPI adjustments, compensation growth indices — are computed directly in dashboard SQL queries rather than a separate normalized schema.
|
||||
|
||||
**Rationale:** With a single institution (UD) and a populated DuckDB, the query layer is sufficient for a local prototype. A proper Phase 2 with dbt transformations and a unified analytical schema should be built before:
|
||||
- Expanding to multi-institution comparisons (Phase 4)
|
||||
- Moving to a production React dashboard (Phase 3 full build)
|
||||
- Adding complex cross-source joins that benefit from materialized views
|
||||
|
||||
---
|
||||
|
||||
## Technical Decisions
|
||||
|
||||
| Decision | Choice | Rationale |
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue