Data scientist · London, UK

Sean Pedrick-Case

I build open-source ML and AI tools for resource-constrained organisations. Interested in the safe deployment of evaluation of AI solutions.

Projects

A selection of open-source projects I've recently worked on. Each one is on GitHub, and most have a live demo linked underneath that you can try (on Hugging Face).

Featured

All projects

Blog

Notes and experiments from my projects (for now, articles on the document redaction app, but I will add to this in the near future).

All posts

About

I'm a data scientist working in local government in London. My work sits at the intersection of public services and AI in a resource-constrained environment. In my time in my current role I have created numerous data science tools to help improve services in the borough. These include an app to analyse open-text survey data with LLMs, a tool to link free text address data, and an app to make sensitive documents safe to share through redaction.

For all of these solutions, safety, and reliability are paramount to best serve our residents. For this reason, I have a particular interest in creating highly secure AI solutions with evaluation and monitoring at the forefront of design.

I work on other projects that I can't yet share publicly as a repo, but I have also done a lot of work to create machine learning models within a social care setting to improve the efficiency of repetitive processes in the review process. More details on my other projects can be seen on my LinkedIn profile.

  • Python
  • LLM applications & topic modelling
  • Record linkage & fuzzy matching
  • PII detection & redaction
  • Gradio apps
  • SQL & data pipelines
  • AWS (Bedrock, etc.)
  • Local government / public sector data

Get in touch

The best way to reach me is on LinkedIn: