GOVINEER ESTATE · LAND USE PLANPrepared by Phil Morehead, candidate for Head of Product
Sheet 1of 6
2,300+ local governments · 43 states
Oct 2026
Drawn from public facts. Parcel calls are my outside read. Demo records are synthetic.
Rev. A
Govineer estate land use plan A town map where each Govineer product area is a zoned parcel. Streets are the shared platform. Hatching marks state and local rules. A dashed teal overlay marks where I'd put AI in front of a clerk first.

Swipe to pan the map. Tap a lot to read it.

Sheet 2 · Parcel U-1

Billing exceptions desk

The first AI a clerk should trust is one that catches the bad bill before a resident does. Work the queue below. The AI reads, explains, drafts and waits. The clerk decides, and every call lands in the audit log.

Town of Ashgrove Bend · water utility
October cycle · 412 reads
5 exceptions held for review
Synthetic town and data

    Evals before it ships

    214exceptions replayed
    92%matched the clerk
    11false holds
    3misses, listed
    • Missed a slow leak that stayed under the 3x flag for two cycles. Fix: trend rule, not just a threshold.
    • Missed a meter swap that reset the read to zero. Fix: pull the work order history in.
    • Released a bill the clerk held for a dispute that lived in a phone call. Fix: clerk notes become an input.

    Replayed against 12 synthetic cycles. The other 4 cases the AI handed to a person. Ship gate: 90% match on holds, zero actions without a clerk, 10% of releases sampled weekly.

    What the AI sees

    Account number, rate class, 13 cycles of reads, service status and the town's own ordinance text. No names, no addresses, no payment data.

    It runs inside the customer's tenant. Every run is logged with its inputs and output, so an auditor can replay any bill.

    The runs on this page are precomputed. No live model calls.

    Sheet 3

    Rezoning rules

    A rollup doesn't get one platform by decree. It gets there one parcel at a time, with rules everyone can see. These are the five I'd run every portfolio decision through.

    Zoning callThe testOn this map
    Share itThree or more product lines rebuild it, and no state rule makes it differ. Build it once, behind an API that legacy and new apps both call.Payments rail, resident account, identity, notifications, audit log.
    Keep it localA state law, an audit rule or a customer commitment makes it differ. Ship it as configuration on the shared base, never as a fork.Withholding, shutoff notices, fund structure, tax rules, permit codes.
    Overlay AIThe clerk does it often, the right answer is checkable, and a wrong one gets caught before a resident or a ledger sees it.My proposal: billing exceptions first, then AP coding and permit intake as pilots.
    Retire itTwo products do the same job for the same buyer, and the customer can move with no lost data and no lost state feature. Proven on real migrations first.Decided with Govineer's data, not drawn from the outside.
    Hold the lineAI that would act on a resident or post to the ledger with no person in the loop. Not yet, and maybe not ever.Auto shutoffs. Auto posting. Auto denials.

    First three rezonings I'd propose

    1. One payments rail and one resident account across every product. Residents feel it first, and it carries pricing upside.
    2. Shared identity and audit log, so AI runs and clerk actions replay the same way in every product.
    3. Billing exceptions with clerk review, piloted in a handful of towns, with evals published to the team every week.
    Sheet 4

    First 90 days

    Survey before you plat. Plat before you build. Each phase ends in something the team can hold me to.

    Survey

    Days 1 to 30
    • Sit with clerks and finance officers in at least three states, across every product line.
    • Meet each acquired team on their turf. Learn what they're proud of before asking them to change.
    • Read support tickets, churn notes and lost deals by product.
    • Baseline every product: activation, retention, support load, margin.
    You getThe real land use map. This one, corrected by people who know the ground.

    Plat

    Days 31 to 60
    • Make the share, keep local and retire calls, each with an owner and a reason.
    • Give every product one outcome metric its team owns.
    • Bring pricing and packaging hypotheses for the shared layer.
    • Set the PM hiring plan and the bar for product reviews.
    You getAn investment memo with explicit stops, not just starts.

    Permit

    Days 61 to 90
    • Pilot billing exceptions with clerk review in a small set of towns.
    • Stand up the eval suite and the weekly review with engineering.
    • Publish the first shared service API contract with Platform.
    • Run the product operating rhythm across every team.
    You getPilot results with evals attached. Real clerks, real cycles, not a demo.
    Sheet 5

    Week one questions

    What I'd ask Manav Thiara and the team first. The answers redraw every sheet above.

    1. Which product does a customer feel first when we get it wrong?That's where trust gets spent, and where AI ships last.
    2. Where do the acquired teams disagree about the platform today, and who breaks the tie?Rezoning stalls on unclear decision rights, not on code.
    3. What does the AI operating system need to show by the next board cycle?So the first pilot proves the thing the investment is betting on.
    4. Which state rules cost the most engineering time per customer?That's the configuration layer to build before anything else goes shared.
    5. How do we measure retention and expansion per product now, and which numbers do you trust?I'd rather know the gaps on day one than find them in a board deck.
    6. What did a customer stop paying for last year, and why?Retirement calls start with what customers already decided.
    7. What do you want off your plate by day 90?The Head of Product job is partly giving the CPTO time back.
    Sheet 6

    The surveyor

    Sixteen years in product. I've built a product org from scratch and shipped on top of legacy systems. I've set the success metric for every launch I've touched in the last five years.

    1. 2021 to 2025Director of Product, Launch by NTT Data. Led a 60+ person product org with 15 direct reports, 10 of them product leaders. Built NTT Data's first product org and moved about 45 client accounts from a BA and PM model to product, outcomes and design in 12 months. Launches got about 50% faster, with fewer escalations and better retention. The model has since scaled to hundreds of PMs.
    2. Legacy plus APIHCA hospital intake app. Shipped on Epic and Meditech data over FHIR and HL7. Owned the device rollout, MDM and field reliability. Same shape as Govineer: new experiences on systems you can't rip out.
    3. 2025 to nowPrincipal Product Strategist and AI Ambassador, Sparq. Led the Product Strategy Practice. Now I lead AI offers, price deals and build the tools our teams use to ship with AI, like the prototype on this page.
    4. 2024 to nowCo-founder and CPO, NutrAssist. A family nutrition app in beta with about 250 users. I own positioning, pricing and what ships.
    5. 2017 to 2019Business Process Analyst and Systems Lead, State of Ohio DODD. Led identity and access (OAuth, SAML) for HIPAA sensitive apps serving statewide care programs. I've been on the government side of the software.
    6. FoundationB.S. in Health Administration, Ohio University. I started in regulated, public facing service work and stayed close to it.

    AI credentials

    • Anthropic
    • Claude Certified Architect, Professional874
    • Claude Certified Architect, Foundations851
    • Claude Certified Developer, Foundations911
    • Claude Partner Badge, Claude Code2026
    • AWS
    • AWS Certified AI Business Strategist2026
    • Hands on: Agentic AI, MLOps, Agent Identity, Databases for AI, Serverless5
    • Google Cloud
    • Gemini Enterprise skill badges, Deployment and Agent Development tracks6
    Verify on Credly
    What I don't have yet: the vendor side of GovTech.

    I've worked inside a state agency's systems. I haven't sold to or shipped for local governments. I close gaps like that by sitting next to the people who do the work, so expect me with utility clerks in my first month. You'll see it on the map.

    Let's walk the map together.

    Thirty minutes. Bring the parcels I got wrong. That's the conversation I want.

    Email Phil LinkedIn Resume