CHARLIE HOWMAN

Software Engineer.

Computational chemist turned software engineer, from predicting atomic forces with neural networks to shipping production RAG pipelines at PwC.

EXPLORE MY WORK

MY
BACKGROUND

Experience

Software Engineer, Technical Delivery @ PwC UK

Sep 2023 – Present

Progressed from PwC's Technology Graduate Programme into a Software Engineering role within the Technical Delivery team, building full-stack applications spanning frontend, backend, database design, and AI-driven workflows. Led development of TestGenPro, PwC's generative AI test automation tool, architecting a Node.js/SQL Server backend behind a Next.js frontend. Currently building a Risk Platform comprising four modules for managing risk and control matrices, working across API design, database architecture, and UI to deliver an end-to-end user experience. Designed and shipped production RAG pipelines (chunking, embedding, and retrieval via Azure AI Search) powering a live chatbot for PwC's Risk division, and integrated LLMs into client-facing ESG and supply chain prototypes.

Research Placement @ Manchester Institute of Biotechnology

Jul 2022 – Sep 2022

Built and optimised batch simulation workflows in Python for molecular systems, applying computational models to study biodegradation of environmental toxins.

Skills

Languages & AI/ML

Python
LangChain / LangGraph
RAG Pipelines & Agents
Azure AI Search

Frameworks & Data

SQL
Next.js / Node.js
React
Databricks / PySpark

Cloud & Infra

Docker & CI/CD
Azure

Education

MChem, Master of Chemistry @ University of Manchester

2018 – 2023

First Class Honours. Focus on computational chemistry and artificial neural networks. Final year project developed an ANN using TensorFlow to predict atomic forces in sugar molecule systems, optimising simulation accuracy.

Interactive: ethylene glycol conformational dynamics — drag to rotate, scroll to zoom.

WHAT I'VE
WORKED ON

Some of the things I've built while learning, experimenting, and bringing ideas to life.

TestGenPro

PwC's generative AI test automation tool, transforms user stories into structured test cases and steps, with a Node.js/SQL Server backend behind a Next.js frontend.

RAG Chatbot for PwC Risk

End-to-end pipeline chunking, cleaning, and embedding large document corpora into Azure AI Search, powering natural language Q&A over structured and unstructured data for a production chatbot.

Neural network visualization of atomic force predictions

ANN Force Prediction for Sugar Molecules

Final year MChem project. Built an artificial neural network with TensorFlow to predict atomic forces in sugar molecule systems, improving molecular simulation accuracy.

Coming Soon

Stay tuned for exciting new projects currently in the works!

MY
HOMELAB

A small self hosted server I run at home so I have a hands-on way to learn networking, containers, and infrastructure outside of work.

Home Router

Network Switch

Dell OptiPlex 7020 16GB RAM · Docker

Uptime Kuma Pi-hole Portainer

The Build

An old Dell OptiPlex 7020 (16GB RAM) repurposed into a home server, connected via a dedicated network switch off my home router. Everything runs as Docker containers, managed through Portainer, so services are easy to add, update, and tear down without touching the host directly.

What Do I Host On My Server?

This portfolio site — you're looking at it, live
Uptime Kuma — uptime monitoring
Pi-hole — network-wide ad/tracker blocking
Portainer — container management
Docker

Live Status

Real-time uptime monitoring for the services above, via a public Uptime Kuma status page.

Is this site live?