From the field book to the data-use register

Mandate

Set the rules for how customer, location and telemetry data may be used as the company acquires products and introduces machine-learning features.

Outcomes

• Introduced a separate consent for model training; 126 customers had signed by June 2026
• Migrated 3.1 million acquired records with datum and confidence lineage
• Narrowed a benchmark product to regional aggregates with a twelve-municipality minimum
• Added data-lineage review to acquisition due diligence

Decisions in context

The chief data officer role was created in 2021, when Halvercroft's investment brought an acquisition programme and a board that wanted data risk reported as carefully as financial risk. My work divides into rules and reconciliation. The rules decide which uses of customer data each contract permits, and they now govern model training, analytics and aggregated products separately. The reconciliation work brings acquired products into one consistent account of position, schema and ownership. I approve data uses but not product strategy or commercial terms. I attend board meetings when data matters are on the agenda and report to the audit and risk committee each quarter on incidents, access reviews and open consent questions. The three cases on this site all come from this role, and each involved a colleague whose objection to my recommendation was reasonable.

Reason for the move

Current role; seeking a first independent board seat alongside it.

Mandate

Build a product analytics capability and a single customer-data inventory that public-sector buyers would accept.

Outcomes

• Launched event-based product telemetry with a published event list and account-level switch
• Moved Canadian customer data into Canadian hosting regions by late 2019
• Created the first register of data permissions across 1,900 customer contracts
• Shortened average municipal privacy review from eleven weeks to five

Decisions in context

I joined Pellarwick because, as a customer, I had asked questions the company could not yet answer. The founders were candid about that. My first year was spent listing every dataset we held, where it was stored and which contract governed it, a task that took far longer than anyone expected because older agreements had been negotiated customer by customer. I then built a product analytics programme around events rather than individuals, so that municipal privacy officers could approve it. The move of Canadian customer data into Canadian hosting regions took eighteen months and required re-testing every integration. I reported to the chief technology officer until 2019 and then directly to the chief executive. The growth investment in 2020 brought board-level scrutiny of data risk, and with it a request to formalise my responsibilities.

Reason for the move

Appointed chief data officer when the company's new investor and board created the role.

Mandate

Make the network records reliable enough for planning, outage response and excavation requests.

Outcomes

• Introduced positional-confidence classes for 210,000 network assets
• Reduced the backlog of unrecorded as-built changes from fourteen months to three
• Set data-sharing terms for the company's mapping and analytics suppliers

Decisions in context

Keswold's records described poles, transformers and underground cable across a large rural service territory, much of it built before digital mapping. Field crews trusted the map until it was wrong once, and after that they trusted nothing. I introduced confidence classes so that a crew could see whether a cable position came from a survey, an as-built sketch or a scaled paper plan. My team cleared a fourteen-month backlog of construction changes by working with construction supervisors to capture as-builts at job closure. I also wrote the data-sharing terms for our software suppliers, including Pellarwick, after discovering that a field application was reporting our crews' locations to a vendor dashboard. Those terms were the reason Pellarwick's founders called me. I learnt there how a supplier's quiet data habit can damage an operating relationship that is otherwise sound.

Reason for the move

Recruited by Pellarwick, one of Keswold's software suppliers, to lead its data and analytics work.

Mandate

Maintain the county's parcel, road and drainage records and manage requests to share them.

Outcomes

• Rebuilt the parcel fabric against current survey control across 2,900 square kilometres
• Wrote the county's first data-sharing protocol for consultants and developers
• Digitised 11,000 historical drainage and road plans with source annotations

Decisions in context

The county was the first place I held records on behalf of the people they described. Ratepayers expected their property information to serve assessment and planning, not to appear in a consultant's marketing material. When an engineering firm asked for the full drainage inventory, including notes about private culverts, I wrote a sharing protocol that set out purpose, permitted use and return or deletion. Councillors approved it after two meetings. The technical work was equally formative. Our parcel map had been assembled from plans of different ages, and rebuilding it against current survey control moved some boundaries by several metres on screen without changing anyone's legal boundary. I spent many evenings explaining that difference at public counters. My director, Ruthanne Pelchat, taught me that a careful explanation offered before a ratepayer notices a change is worth more than a perfect correction offered afterwards.

Reason for the move

Moved to a utility to work on infrastructure records with greater operational consequence.

Mandate

Turn field observations and historical plans into drawings and coordinate files ready for professional review.

Outcomes

• Standardised the firm's coordinate-file conventions across three survey crews
• Created a reference library of historical plans with datum notes
• Cut average drafting turnaround from nine working days to six

Decisions in context

Moving from the field to the office showed me what happened to observations after the crew went home. I prepared drawings for commissioned land surveyors to review and sign, and I had no authority over their professional conclusions. What I could do was make discrepancies visible. When a new survey disagreed with a plan from the 1950s, I documented the reference system and evidence behind each rather than adjusting one to fit the other. I built a small library of historical plans annotated with their datums, which the partners used for years afterwards. I also learnt how clients treated the data we delivered. Developers and engineers reused our coordinate files for purposes we had not anticipated, sometimes reasonably and sometimes not. That observation, more than any formal training, led me towards questions about permitted use.

Reason for the move

Joined a municipality to manage land records directly rather than prepare them for clients.

Mandate

Collect accurate field observations for legal, topographic and construction surveys.

Outcomes

• Led more than 400 field days as crew chief
• Introduced a daily closure check before leaving each site
• Trained four junior technicians in instrument and safety practice

Decisions in context

My working life began outdoors with a total station, a prism pole and a party chief who would not accept a traverse that failed to close. We surveyed well sites, subdivisions and road widenings, often in weather that made careful work difficult. I learnt that a measurement is a claim about the world that someone else will rely on, sometimes decades later. As crew chief from 2000, I was responsible for the quality of each day's observations and for a junior colleague's safety on roadsides and rural leases. I introduced a habit of checking closure before leaving a site, which saved several return trips. The work also taught me humility about records. The evidence we found in the ground frequently contradicted the plan we had brought with us, and the plan was not always the one in error.

Reason for the move

Moved into the office to work on data preparation and drafting.