cv
Basics
| Name | Aliaksei Pilko |
| Label | Senior Data Scientist & Simulation Engineer |
| jobs@aliakseipilko.com |
Work
-
2024.07 - Present Senior Analyst (Data Science & Simulation Engineer)
AiQ Consulting
Technical lead for next-generation airport simulation platform, responsible for architecture, backend, frontend, and DevOps.
- Led development of integrated airport simulation replacing unstable legacy tooling; stack: Python, Django, React, CesiumJS, Docker, GCP.
- Built CI/CD pipelines (GitHub Actions) and observability stack (Grafana, Prometheus, Sentry).
- Introduced Dagster-based data pipelines for company-wide data lake; improved project data reuse and reduced per-project setup time by days.
- Mentored two junior engineers from data backgrounds to full-stack developers, reducing key-person risk.
- Technical lead on £180k Heathrow optimization project: developed and presented actionable & evidence-based recommendations reducing Polars, DuckDB, Dagster.
-
2020.05 - 2024.06 Researcher / PhD Candidate
University of Southampton
Research in UAS operational risk analysis and airspace simulation, combining probabilistic modeling, HPC simulation, and real-world data integration.
- Developed probabilistic and agent-based UAS risk models using Monte Carlo and rare-event simulation. Productionized into full stack web app and used in flight trials.
- Architected 500k-records/day pipeline for novel sensor data (cameras, radars, passive RF) with 99 % uptime (Kafka, Docker, Go, TimescaleDB, Grafana) for DfT Innovation project.
- Developed high-performance C++ risk and deconfliction libraries with Python bindings (Eigen, GEOS, OpenMP). Integrated into web app for field trials.
- Led development of multi-objective logistics optimization system combining research from multiple universities into deployed web app (Flask, React, Gurobi, ipopt).
- Automated classification of 60 k medical-goods PDFs via web-scraping + Gemini/Vertex AI, leading to follow on grant funding.
-
2019.05 - 2019.09 Machine Learning Intern
Tekever
Developed ML models for maritime surveillance UAVs.
- Enabled autonomous monitoring of 10x larger maritime areas through ML trajectory prediction and anomaly detection of ship movements.
- Integrated ML models into UAV GCS systems for field deployment in bandwidth-limited environments.