Benyamin Ebrahimpour, PhD, AFHEA

AI/ML Engineer · Battery Modelling · Bayesian Inference

Southsea, Portsmouth, UK · Benyamin.ebrahimpour@port.ac.uk · UK Driving License (Automatic)

About

I’m a highly motivated AI/ML engineer with a multidisciplinary background in machine learning, data science, and mathematical modelling. I hold a PhD in Mathematics and build neural network surrogate models and Bayesian inference pipelines for lithium‑ion battery systems. I enjoy shipping clean, production‑ready code and collaborating in cross‑functional teams.

Currently: Finalist (3rd place runner‑up) at the IGNITION Responsible AI Startup Hackathon for an AI‑based battery health monitoring solution, awarded $2,000 in AWS credits.

Core Stack

Python, R, SQL, PyTorch, NumPy/SciPy, GPyTorch, pandas, scikit‑learn, Tableau, Power BI, Excel, JavaScript, HTML/CSS, MATLAB, Maple, Fortran, C, PyBaMM, PyBOP.

Also: PSO, MCMC/Metropolis, Bayesian UQ, Gaussian Processes, physics‑informed ML.

Domains

Battery modelling (ECM & physics‑based), parameter estimation, uncertainty quantification, energy systems analytics, data visualisation, teaching & public engagement.

AI/ML Batteries Bayesian Data Viz

Experience

Data Analyst Intern · Victoria Solution (2025)

London, UK

  • Advanced analysis with Python/R/SQL; built interactive dashboards in Tableau & Power BI.
  • Sourced, cleaned and structured large datasets; presented insights to stakeholders.
  • Collaborated with industry mentors and data teams; delivered findings in feedback meetings.

Research Assistant · University of Portsmouth (2024–2025)

  • PSO + Bayesian inference with neural surrogates for parameter estimation (custom Python).
  • Applied PSO to inverse problems; MLE & MCMC (Metropolis) for UQ.
  • Built Li‑ion models with PyBaMM/PyBOP; integrated experiments and numerics in Git‑driven workflows.

PhD Researcher · Battery Modelling & Surrogates (2021–2024)

  • Developed scalable neural surrogate models for Li‑ion batteries enabling ultra‑fast Bayesian inference.
  • Physics‑informed & supervised learning pipelines; Gaussian Processes with GPyTorch.
  • Reduced‑order blended‑electrode model (MATLAB finite volume) for control‑oriented simulation.
  • Collaborations with The Faraday Institution; national conferences and publications.

Selected Projects

BatteryGuard AI (Demo)

AI‑powered health check for second‑life Li‑ion batteries. Includes model parameter inference via PSO + neural surrogates and an educational microsite.

Open demo section GitHub

EV Population Dashboard (Tableau)

Interactive dashboard exploring EV population in Washington State — trends by brand, geography and registration year based on open data.

View Tableau

BatteryGuard AI — Interactive Mini Demo

For GitHub Pages, backend calls are disabled. This simulated demo previews the UI flow only.

Upload CSV (simulated)

What it does

  • Fits a Thévenin ECM to current/voltage data.
  • Uses a trained NN surrogate + PSO search.
  • Reports R0, R1, C1 and notes on fit quality.

Publications (selected)

See Google Scholar

Teaching

Awards & Memberships

PhD, Mathematics — University of Portsmouth (2021–2025)

Dissertation: Neural Network Surrogates for ultra‑fast Inference in Li‑ion battery models.

MSc, Mechanical Engineering — Sharif University of Technology (2019–2021)

Top 2% on entry; GPA 18.53/20. Dissertation on solar desalination systems.

BSc, Mechanical Engineering — Babol Noshirvani Univ. of Technology (2014–2018)

Top 8% (3rd Top Student). Dissertation on bio‑inspired underwater robots.

Contact

Email Benyamin.ebrahimpour@port.ac.uk

Location Southsea, Portsmouth, UK

Links LinkedIn · Google Scholar · GitHub