Proof of Work
    Proof of Work

    Built and governed, not just advised

    Few advisors have built what they advise on. These are systems I designed and run myself, and what they taught me about governance shapes every engagement.

    Case 01

    ProjectView PM: an AI PMO office with a human in the loop

    Problem
    Weekly portfolio and status reporting is slow, manual and inconsistent.
    What I built
    A staged pipeline of seven tool-using AI agents. A Program Coordinator; then Risk & Dependency, Resource and Finance agents working in parallel; a PMO Analyst that scores health (Green / Amber / Red) and KPIs; an Executive Communications Assistant that assembles one report and a steering deck; and a Stakeholder Communicator. Agents share findings through a common findings board and ground their work on the app's own data: budget, tasks, timeline, meeting notes and calendar.
    Governance pattern
    Every consequential action (sending email, creating tasks, editing timelines, updating resources) is queued to a human inbox to review, edit, reject or approve, and nothing executes without approval. Every run is logged with the tools used and tokens spent, for full traceability. It runs weekly on a schedule or on demand.
    Result
    About 40% less manual reporting effort in my own use.

    What it taught me

    "Put human approval at the point of consequence, not everywhere. That's what makes agents safe enough to use."

    Case 02

    PMO Gateway: stage-gate governance with AI that drafts but never approves

    Problem
    PMOs run stage gates on documents that arrive late, incomplete or inconsistent, and approvals live in email. When an auditor or executive asks whether a project cleared its gate with the right approved documents, nobody can prove it.
    What I built
    A branded workspace where each company defines its own gates, phases, document catalog and templates, with no fixed PMO framework. A project's tailoring profile (governance level, vendors, agile delivery, regulated data) decides which documents each gate requires. Built-in AI helps project managers draft and revise documents, interviews them to fill gaps, ingests meeting notes, imports existing material, checks consistency across documents and summarizes committee rejections. Committees review and sign off gates; outputs export to PDF and DOCX; a PMO dashboard shows governance and portfolio views.
    Governance pattern
    AI drafts; people approve. Only a committee approval satisfies a gate document. The project's tailoring profile locks at first submission or gate sign-off, so requirements can't be quietly lowered; changes go through a logged change request. Every document version keeps its sources and change reason. AI use is metered per organization with a monthly allowance and logged by action and tokens. Imported legacy projects are attested, never shown as back-dated approvals.
    Result
    In production as a multi-company platform with organization workspaces, paid author seats and free trials.

    What it taught me

    "Let AI write the first draft, never the approval. Governance lives in who signs, what's locked, and what's logged."

    Case 03

    Legacy Shield AI: guided estate planning with AI that knows where to stop

    Problem
    Most families never finish an estate plan. Attorney-led plans are costly, and do-it-yourself forms ignore each state's signing and witness rules.
    What I built
    A self-serve platform with a conversational AI consultant that guides intake and produces legally formatted documents for all 50 states plus DC, applying each state's signing, witness and notary rules: will, revocable living trust, pour-over will, powers of attorney, healthcare directive, living will, HIPAA authorization, deed transfer, legacy letter and pet care plan. It adds a step-by-step trust funding workflow, an encrypted Digital Vault, household members and trusted-contact sharing. Families build for free and pay only to download finished documents.
    Governance pattern
    Clear boundaries and a human escalation path. The platform states plainly that it is not a law firm and does not give legal advice. Optional attorney review and notary services hand off to licensed professionals. Credentials in the vault use zero-knowledge, browser-side AES-GCM-256 encryption, and household sharing defaults to view-only.
    Result
    Live, covering all 50 states plus DC.

    What it taught me

    "In a regulated domain, the most important thing an AI can do is know where to stop and hand off to a human."

    Also built

    ResumeRefreshTax-CChipInSympodio

    Want this rigor applied to your AI?

    Book a free scoping call