Product Manager — Data Platform · vacancy
Here's what AI can do for this role — and what still needs a human. Built straight from Kestrel Utilities’ own job advert, running live on my_db.kestrel_demo.accounts — 27,640,145 real rows via MotherDuck (DuckDB). Not a slide about AI. The job, getting done.
Every line on the left is lifted from Kestrel Utilities’ actual job ad. If a card lacks a harvested JD line, it is omitted. On the right is the AI doing it — with eligible cards running live against the warehouse and offline inspection clearly labelled in the workspace.
“Own the strategy, vision, and roadmap for the PDL as a governed data product — engaging external and internal consumers, the Analytics Chapter, and engineering leadership to define, validate, and continuously iterate the direction.”
How fresh is the meter-read pipeline — mean days since last read by meter technology?
bar chart“Own the commercial narrative for the PDL as an external product offer — defining what Kestrel Utilities’ data capability means to an energy retailer’s analytical team.”
How complete is the monthly consumption pipeline — accounts with ≥10 months of records vs sparse accounts?
bar chart“Make evidence-based prioritisation decisions, estimating the effort and value of roadmap items, using AI-powered estimation tools to improve accuracy, and discussing trade-offs transparently with stakeholders.”
What percentage of active accounts have a known EPC band — how complete is the enrichment pipeline?
kpi“Define and track key product outcomes — implementing dashboards for real-time performance visibility, and using data to drive continuous improvement in quality and user experience.”
Which accounts are high-consumption outliers — est_annual_kwh > 20,000 kWh flagged for data quality review?
deviation“Represent users confidently in internal discussions, advocating for research across all user types and leveraging AI-powered tools to accelerate insight generation at scale.”
What is the monthly data volume trend — total kWh ingested and record count by month across the platform?
bar chartThe honest other half. AI does the analysis; a person owns the decision — especially where regulation, fairness and accountability bite.
A plain-English question — the same one the job ad describes — is translated to SQL by the agentic backend.
Curated cards run server-side against MotherDuck when eligible. The workspace separately labels any local inspection path.
Runs against my_db.kestrel_demo.accounts (27,640,145 rows declared by the manifest). No synthetic numbers.
Each figure carries a falsifier — recomputed from the result set, not a stored number, so it can't quietly drift.
It's the role getting done: curated questions run live server-side against the warehouse; local inspection is labelled inside the workspace.
Open the live workspace →Provenance. Representative Kestrel Utilities-style operational dataset (480 accounts · 728 meters · 5,676 monthly consumption records). Schema mirrors my_db.kestrel_demo.accounts/meters/consumption_monthly. Seed 20260609 — reproducible. No real Kestrel Utilities or customer data. Live server-side path: my_db.kestrel_demo.accounts (27.6M rows). Dormant until operator provisions MOTHERDUCK_TOKEN.
itsorted — I took Kestrel Utilities’ job ads and didn't write a report on what AI could do. I built it. Get the rest sorted →
I'm trained on this proof and the real Kestrel Utilities: the Meridian meter-to-cash platform (seven modules), the move under Helios Group in 2024, 7M+ energy accounts migrated for suppliers like Northwind Energy and Pennine Gas & Power, and the Ofgem framing. Ask me how the Data Analyst function changes shape, or which open roles map to which Meridian module.