Skip to content
View AlexBatrakov's full-sized avatar
😄
😄
3 followers · 32 following

Highlights

  • Pro

Block or report AlexBatrakov

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
AlexBatrakov/README.md

Hi there 👋

Alexander Batrakov

Data Scientist / Analytical Software Engineer PhD candidate in Physics with 5+ years of experience building reproducible analytical workflows, statistical modeling pipelines, and data-intensive software systems.

Interested in:

  • Data Science
  • Machine Learning
  • Analytics Engineering
  • AI Tooling
  • Scientific and Analytical Software

Core Areas

  • Data Science & Machine Learning
  • Statistical Modeling & Time Series
  • Experimentation & Evaluation Frameworks
  • Analytics Engineering
  • Reproducible Workflows
  • Python, SQL, Julia
  • Docker, CI/CD, Cloud Deployment

Featured Projects

Wearable Analytics

Privacy-first analytics workflow on real Garmin wearable data. Python, SQL, DuckDB, scikit-learn, statistical validation, time-aware modeling and dashboard-oriented analytics.

PyTorch Pets Classifier

End-to-end computer vision project with reproducible experiments, robust model selection, calibration analysis, FastAPI serving, Docker and Azure deployment.

Solo Wargame AI

Simulation and experimentation platform for a stochastic tactical decision environment with exact references, benchmark evaluation, heuristic agents and deterministic replay.

GravityToolsNext

Julia framework for reproducible computational experiments, statistical diagnostics and uncertainty-aware analytical workflows.

Evalynx

Backend control plane for reproducible computational runs with FastAPI, PostgreSQL, Redis + RQ, and a real external runner integration.

Tech stack

Languages: Python · SQL · Julia · C/C++

ML & Data: pandas · NumPy · SciPy · scikit-learn · PyTorch

Engineering: Docker · GitHub Actions · FastAPI · Linux

Analytics: DuckDB · PostgreSQL · Statistical Testing

Selected Highlights

  • First-author peer-reviewed publication (A&A, 2024)
  • Built analytics workflows on 23-year irregular time series
  • Developed reproducible experiment frameworks used in PhD research
  • Deployed ML services using FastAPI, Docker and Azure

Pinned Loading

  1. pytorch-pets-classifier pytorch-pets-classifier Public

    PyTorch Oxford-IIIT Pets classifier with reproducible experiment tracking, seed-sweep model selection, error analysis, and a live FastAPI deployment on Azure Container Apps.

    Python

  2. wearable-analytics wearable-analytics Public

    Privacy-first Garmin wearable analytics: JSON-to-Parquet pipeline, sanitization & quality gating, SQL layer (DuckDB + PostgreSQL showcase), EDA, statistical validation, and Stage 3 next-night recov…

    Python

  3. solo-wargame-ai solo-wargame-ai Public

    Rule-faithful Python simulator, search baselines, and RL experiments for a stochastic solo hex-based tactical wargame.

    Python 1

  4. evalynx evalynx Public

    Backend control plane for reproducible computational runs with FastAPI, PostgreSQL, Redis + RQ, and a real external runner integration.

    Python

  5. GravityToolsNext.jl GravityToolsNext.jl Public

    Julia orchestrator for reproducible TEMPO/TEMPO2 workflows + pure-Julia adaptive grids

    Julia

  6. StructureSolver.jl StructureSolver.jl Public

    Julia package for relativistic neutron-star structure in DEF scalar–tensor gravity (TOV + post-processing).

    Julia