Data & AI
What I do
I help small and mid-size teams clean up their data, ship useful analytics, and add AI where it actually helps. My background is in data engineering and analytics, and I've been shifting towards AI-augmented workflows: agents that support reporting, recruiting, and operational decision-making. Instead of big platform projects, I focus on concrete outcomes a team can start using in days or weeks, not months.
Who I work with
I work best with companies still building their data foundations or modernising from spreadsheets and ad-hoc reports. This often means founders, operations managers, and small data teams who need someone to design the pipeline, automate key flows, and document everything clearly. I'm open to remote or partly-remote collaboration within European time zones, and comfortable as a consultant, contractor, or embedded team member for a few months at a time.
Typical work
Data pipelines
Setting up or improving ETL/ELT pipelines, defining data models, and integrating multiple sources into a reliable reporting layer.
AI & agent workflows
Integrating LLMs into existing processes — Q&A bots, document extraction, automated summaries, or decision-support agents.
Analytics & dashboards
Building reporting layers in BigQuery, Databricks, or MSSQL and delivering them via Looker Studio, RShiny, or custom UIs.
Data quality
Automated validation frameworks that catch anomalies before they reach stakeholders — rule-based or ML-assisted.
Selected work
Insurance Data Platform Migration & Quality Engineering
Zurich Insurance Company Ltd · 2016 – 2026Supported the migration of a global insurance reporting platform from Palantir Foundry to Azure Databricks, then maintained and extended the resulting lakehouse architecture. Built PySpark-based data quality frameworks that validated multinational actuarial datasets across pipeline runs, and delivered RShiny dashboards consumed by business stakeholders across multiple countries.
Reliable reporting layer for actuarial and underwriting teams across 10+ countries, with automated anomaly detection replacing manual checks.
Marketing Analytics Pipeline on GCP
Chosen Data (freelance) · 2022 – presentBuilt end-to-end marketing analytics pipelines ingesting event and campaign data from Google Analytics 4 and Google Ads into BigQuery. Applied SQL transformations and scheduled queries to produce clean reporting layers, delivered as Looker Studio dashboards for campaign performance monitoring and client decision-making.
Clients moved from manual spreadsheet exports to automated, always-current dashboards — decisions based on data updated daily, not weekly.
Urban Parking Demand Forecasting
GoSpace Labs · 2024 – 2025Developed Python-based predictive models for urban parking demand using scikit-learn and pandas. Covered the full ML cycle: feature engineering on real-world infrastructure datasets, model training and evaluation, and productionising the output as scheduled processing jobs.
Research-stage models turned into reproducible, production-ready pipelines with Git-based versioning — ready for integration into live operations.
This portfolio site (radkoseno.eu)
Personal project · 2026Built this site as a working example of AI-augmented engineering: a Next.js 16 app with Tailwind CSS, deployed on Vercel with auto-deploy from GitHub. A recruiter Q&A bot (in progress) will let visitors ask questions about availability and experience, answered by an LLM grounded in a structured profile JSON.
A portfolio that demonstrates the approach, not just describes it — the bot itself is a live example of practical AI integration.
Have a data or AI problem you'd like to talk through?