Product Owner, Data Sourcing

About us

Dashmote builds AI-native, agentic data products for both Fortune 500 brands (Coca-Cola, Heineken, Red Bull) and growth-oriented businesses. With offices in Amsterdam, New York, and Shanghai, we run two products: Location Intelligence (120+ data points per outlet across 30+ countries) and Food Delivery for optimising Sales & Marketing. We pair both with AI agents that work alongside our clients’ commercial, sales, and insights teams to find, score, and convert the right opportunities. As we scale up our multi-year global data pipeline, our Shanghai sourcing team remains the core engine behind every data product we ship.

The Role & Responsibilities

Your Mission

This role is a rare combination: a technical engineer, a product owner, and a business thinker in one person. Our Shanghai sourcing team runs large-scale web scraping across food delivery platforms, maps, and social media in 30+ countries, including serious anti-bot challenges. The team solves hard technical problems well; you bring the structure and judgment around it.

Your mission is to make sourcing delivery predictable: risks surfaced early, data issues caught before release, and stakeholders always knowing the current scope, quality, and timing.

What You Will Own

1. Technical Depth (Master the System & Its Logic):
You write Python, understand proxies and anti-scraping, and master the difficult logic behind every delivered number deeply enough to challenge the team: Finder completeness (grid coverage, full-grid vs hash-only refresh, and what it does to outlet counts), closed-outlet detection, new-outlet flags between refreshes, deduplication, and refresh/optimization strategy per source (frequency, incremental vs full crawls, request budgets). You document this logic and progressively productize it into repeatable rules per source.

2. The Product Owner (Own Planning, Quality, & Delivery):
One source-of-truth backlog and delivery tracker for all sourcing work. You run sprint planning with estimation: every task has an owner, an estimate, and a committed ETA. You own intake discipline (every request enters with a business owner, use case, acceptance criteria, and priority; urgent work displaces something explicitly), and you set sprint and monthly goals ranked by business value. You define measurable quality thresholds per source (coverage vs expected universe, closed-outlet rate, duplicate rate, freshness, field completeness), own the QA gates that run before any delivery leaves the team, and the monitoring on continuous pipelines (freshness, volume, last-green-run) so a stalled pipeline is detected in hours, not weeks.

3. The Business Thinker (Own Communication & Impact):
You are the single point of delivery communication: a weekly delivery update to the Amsterdam product and client teams on a fixed cadence (shipped, in flight, at risk), fast incident communication when a feed breaks, and a change protocol: any methodology, scope, or schema change that affects downstream data is impact-analyzed, communicated, and signed off BEFORE cutover. You know which client and deliverables consume every source; before a technical or scope change, you quantify which client-facing numbers move and by how much, and what it does to infra cost. Any coverage scale-up comes with a cost projection first (cost per source, per refresh, compute-hours).

4. The Working Style (AI-First):
You run the team's operations AI-first and set the example: AI agents for status reporting, QA assistance and anomaly detection, documentation, and scraper prototyping. You are accountable for defining where AI output is allowed and where human review is mandatory: AI-written code reaches production only through tests and QA gates. You continuously move manual operational work to agents, so the engineer's time goes to hard sourcing problems.

Authority & Boundaries

• You own WHAT and WHEN: Backlog, priorities