Aliaksei Pilko

prof_pic.jpg

Bristol, UK

I’ve been building simulations and software since 2017, driven by a need to understand and quantify uncertainty in complex systems. Most of my work sits at the intersection of simulation, data and applied ML, all with the end goal of building things that actually get used. This means I have created solutions across the stack, going from low level C++ to deploying full stack web apps.

I have a PhD in Computational Engineering and a BEng in Aerospace Engineering, both from the University of Southampton. I’ve worked across academia and with public sector; consulted for defence and civil aviation; and presented research at conferences from Yokohama to San Diego.

Some of the things I have worked on:

Tech Stack

I’ve had the chance to work across many domains and always picked the best tools for the job. Here’s a list of some of the tools I’ve used:

  • Languages: Python, C++ (with bindings into Python), Typescript
    • Previously: Go, Java, Matlab
  • Web: React, MUI, Cesium, deck.gl, Zustand, ReactFlow, Django
    • Previously: Flask, FastAPI, Gin, Oat++
  • AI/ML/Data: polars, (geo)pandas, duckdb, scikit-learn, LLM APIs (Vertex AI, OpenAI, etc.), Dagster, MLflow
    • Previously: PyTorch, Kafka, Numba, Dask
  • DevOps: Docker, Github Actions (CI/CD), Sentry, Grafana, GCP, AWS, various Linux distros (RHEL, Debian, Arch btw)
    • Previously: Ansible
  • Geospatial: GDAL, GEOS, PROJ in C++; PostGIS

Personal

I like to be in the air as often as I can. I prefer things with wings, but an engine is optional. You’ll probably find me flying and instructing in gliders over Sussex or something heavier and noisier over Oxfordshire.

When the weather isn’t great, I’ll probably be tinkering with my Ender 5 Pro 3D printer, out for a run or a ride or breaking things doing cool stuff on my Homelab whilst trying not to take down the services I actually use (Proxmox, Immich, Jellyfin, ntfy).