Execution roadmap
Learning only counts when it terminates in a working artifact, an explicit proof event, and a next interaction with reality.
| Phase | Domain | Forced output | Done means | Proof |
|---|---|---|---|---|
| Weeks 1–2 | Clay / enrichment | 3–5 real enrichment systems: waterfall, scoring, conditional logic, API calls | Real company data processed; coverage, cost and failures measured | Benchmark + failure/fix ledger + templates |
| Weeks 1–2 | Python / SQL / APIs | Technical reps embedded inside the builds | Retrieve → normalize → store/query → error-handle without tutorial dependence | Code + fixtures + tests + rebuild log |
| Weeks 3–4 | CRM / RevOps | HubSpot/Salesforce sync, field mapping, routing, assignment and lifecycle logic | Records move correctly and deterministically through real CRM state | Sync contract + transition tests + architecture |
| Weeks 3–4 | Outbound | Small real activation experiment | Sends/replies/conversations logged against real targets | Experiment ledger + response evidence |
| Weeks 5–8 | End-to-end GTM | ICP → enrichment → scoring → personalization → CRM → sequence → reply tracking | One complete system runs on real data and produces downstream commercial state | Meetings / qualified replies / pipeline / cost where earned |
| Weeks 9–12 | Optimization + proof | Improve one measured bottleneck and publish 2–3 serious case studies | Baseline and post-state are comparable; claims trace to source evidence | Before/after + public case-study bundle |
Parallel career loop
Applications, recruiter replies, screens and rejection reasons are themselves external evidence. The career pipeline is operated like a GTM system: qualified role → matching proof → application → interaction → learned gap → updated proof.