Skip to content

PRODUCT_BRIEF.md

Latest commit

 

History

History
42 lines (29 loc) · 2.01 KB

PRODUCT_BRIEF.md

File metadata and controls

42 lines (29 loc) · 2.01 KB
Raw

Product Brief — Solo EM Performance Workbench

One-liner

A local desktop workbench where an engineering manager builds evidence-backed reviews and promotion cases against uploaded role frameworks — with AI that drafts from citations, not vibes.

Who it is for

Primary: a single engineering manager running reviews for their directs on Linux, Windows, or macOS.
ICs and peers participate via exported/imported forms, not by logging into the app (v1).

Core loop

upload role R&R → tag people → year-round dossier
  → import self/peer bundles → writing desk (+ AI)
  → next-level / promo packet → share export → history

Pillars

  1. Solo EM workbench — one machine, one trust domain; see TRUST_MODEL.md
  2. Role frameworks — upload R&R docs, structure RoleDefinitions, tag each person; current + next role text is the evaluation basis
  3. Templates as versioned contracts — self/peer/manager/promotion forms; pinned per cycle
  4. Dossier — goals, feedback, achievements, documents, optional meeting summaries
  5. Import-path collaboration — self/peer via bundles; no live multi-user in v1
  6. Writing desk — role-fit + conditional next-level gaps; ≤45 min/person target
  7. AI — in-app Settings keys; grounded drafts; private-tier data opt-in per class
  8. History — timelines and trends on the EM’s workspace

Explicit non-goals (v1)

  • Live IC/peer login, server-enforced peer anonymity, upward n≥3 aggregation
  • Cross-manager live calibration (optional file bundles only)
  • Comp matrices, stack-rank quotas, Slack DM surveillance, HRIS replacement
  • Multi-tenant hosted SaaS (future track)

Build order

Encrypted local foundations → people + role frameworks + backfill → fixed-template cycle + import bundles → writing desk → single-EM calib/consistency → AI → optional integrations.

See DELIVERY.md, LOCAL_STACK.md, FEATURES.md.