Data Scientist
Validio ABStockholm
- Anställning
- Heltid
- Publicerad
- 2 sep. 2026
Få ett CV och ett brev skrivna för det här jobbet
Ladda upp ditt CV. Jobbjagaren hittar det i din erfarenhet som jobbet frågar efter och skriver ett CV och ett personligt brev utifrån det. Det som saknas läggs inte till.
Få hjälp att söka jobbet4 ansökningar med CV och brev ingår gratis. Inget betalkort.
Visa originalannonsOm tjänsten
Validio is the agentic data management platform that automates data quality, lineage, and cataloging — so enterprises can finally trust the data behind their analytics, AI, and reporting. As a Data Scientist on our Algorithm team, you build the models at the heart of the product — the Anomaly Detection that separates good data from bad.
A validator watches a metric over time; an ML model decides, for each new value as it arrives, whether it's anomalous and worth waking someone up for. These models run against tens of thousands of time series at once — unattended, unlabeled, with nobody tuning them per customer. You'll own that end to end: reading a metric from a customer's warehouse, running the algorithm online in our pipeline, and persisting the state it carries between runs, fast and reliably at scale.
What you'll do
- Build the algorithmic core — anomaly detection, plus data profiling, SQL generation, lineage parsing, and change point detection
- Take a method from a research paper to a tested prototype, then to a production model running online in the engine
- Work out why a validator over-alerts on a metric — say one with a strong daily and weekly cycle — and reshape how it forms its expected range
- Tell whether a model change actually improved detection across customers, when there's no labelled set to score against
- Keep a validator's metrics correct and fast in SQL across multiple warehouses
