Evidence before adjectives

What is verified, demonstrated, contextual—and still unproven.

This ledger separates inspectable artifacts from business context and external authority. A live URL proves deployment, not adoption. A repository proves inspectability, not impact. Every material claim states its boundary.

Verified

A public artifact or automated catalog directly supports the claim.

Demonstrated

A runnable implementation supports the capability, with explicit environment limits.

Contextual

A real career or business result is described, but primary records are confidential or external.

Live checks

Server-side probes of VAP, AegisAI, and Enterprise RAG /health plus optional observability/status. Warm budget: 3s applies to health only. Render Free interim: first hit after idle often takes 15–40s.

Probing spine APIs…

Dedicated page: /spine-health

Verified

28 published AI architecture essays

The public Substack RSS feed and local catalog show a sustained series from April 2026 through the current control-plane thesis.

Boundary: Publishing cadence demonstrates sustained authorship, not independent readership, citation, or adoption.

Inspect evidence
Verified

5-spine Principal review path

Home, hire, work panel, and /technical-review lead with AegisAI → VAP → Enterprise RAG → Content Factory → ADR hub.

Boundary: Spine is the recommended review path; catalog demos remain available under Full portfolio.

Inspect evidence
Verified

17 catalog products

12 platform URLs and 5 teaching pattern viewers are catalogued and checked by portfolio CI.

Boundary: Several free-tier APIs can cold-start; catalog live does not imply customer production traffic.

Inspect evidence
Verified

17 catalogued ecosystem repos

Every counted repository has an explicit catalog entry and review link.

Boundary: Repository count measures inspectability, not adoption or engineering impact.

Inspect evidence
Verified

33 architecture decision records

The architecture portfolio exposes numbered ADRs and the local validator checks the count when the sibling repository is present.

Boundary: ADRs document decisions; reviewers should still inspect implementation and tests.

Inspect evidence
Demonstrated

10 golden-eval suite kinds with real CI gates

GER suite kinds include RAG answer, adversarial security, harness QA, repo fix, mission gate, graph HITL, brief gate, triage preference, router_invariant (VAP / AegisAI / FinOps), and collaboration_scorecard (AegisLoop CSS/TUE/hard gates). Selected consumers fail merges on regression.

Boundary: Not every suite currently gates every platform; coverage is stated per repository.

Inspect evidence
Demonstrated

Stranger-replayable spine golden path

Public script + JSON artifact: health → VAP/ERAG (auth-gated honestly) → AegisAI approval_required → FinOps meter. See GOLDEN_PATH.md and docs/artifacts/golden-path/latest.json.

Boundary: Full ask→answer needs VAP_API_KEY/RAG_API_KEY; ACF live publish still requires Clerk.

Inspect evidence
Demonstrated

Acme Support Agent Embed — 12 FDE moves on the spine

ADR-032 wedge: OIDC/SAML ACS, SCIM Users/Groups, HMAC webhooks+DLQ, Slack retry/DLQ, Salesforce Case connector, tenant health+TTFV, FinOps tenant budgets, Stripe test invoice preview, IR playbooks, embed_acme templates + probe_acme_embed_panel.sh. Packaged under /fde Embed lab — code_demonstrated vs operator_wiring honesty. ADR-033 thin geo on routing contract (optional X-Allowed-Regions; not a residency envelope).

Boundary: Stripe is test mode; HubSpot/GWS are adapter-contract only; Postgres RLS is Phase-2 (ADR-026). Strict ERAG needs operator twin if Render Strict host is undeployed.

Inspect evidence
Contextual

10→2 staffing intensity reduction

Enterprise case study describes the workflow, architecture responsibility, governance boundary, and reported operational result.

Boundary: Employer-confidential result; underlying operational records are not public.

Inspect evidence
Contextual

Multi-$M revenue and savings delivered

Enterprise case studies (payments, subscriptions, EDI) describe annualized revenue and savings outcomes under employer delivery programs.

Boundary: Employer-confidential figures; primary financial records are not public artifacts.

Inspect evidence
Contextual

20+ engineers led

Career timeline and LinkedIn history describe staff and platform leadership span across enterprise programs.

Boundary: Headcount claims are career-context; org charts and HR records are not published here.

Inspect evidence
Contextual

19 years of enterprise delivery

Career history is cross-checkable through the public LinkedIn profile and the portfolio career timeline.

Boundary: LinkedIn is the credential source; the portfolio does not independently certify employment.

Inspect evidence

What must be earned outside this website.

These are intentionally not presented as completed. They are the remaining evidence needed for a defensible top-1% authority claim.

  1. 1.Three current recommendations that specifically address AI architecture, governance, or Staff+ technical leadership.
  2. 2.One independently used benchmark, eval dataset, or reference package with published adoption evidence.
  3. 3.One accepted conference talk, podcast, external technical review, or cited architecture article.
  4. 4.Public adoption telemetry: contributors, forks, package pulls, repeat users, or named implementation references.