bergam0t Goto Github PK

I'm currently a trainer on the Health Service Modelling Associates Programme.

I was previously a data analyst, and later data scientist, in a mental health trust in the NHS.

I completed my BA in Biological Sciences at the University of Oxford in 2011-14, and my MSc in Health Data Science at the University of Exeter in 2020-21, with a focus on web app development for healthcare service improvement.

I'm currently working on

• animated outputs for discrete event simulation event logs
• user-friendly visualisation of large graph networks in Streamlit (including adding extra features to the excellent st-cytoscape extension)
• open-source teaching materials on geographical visualisation, location allocation problems, gradient boosting methods, explainable AI, web app development with Streamlit, reproducible outputs with Quarto, forecasting, process mining with pm4py, and more...

My current favourite things to work with are

• Streamlit (particularly serverless deployment with the amazing stlite package)
• Kepler.gl
• NetworkX and Cytoscape.js
• Quarto

I'd love to chat about

• how to use routinely collected data better to improve mental healthcare services
• how to make it easier for analysts and data scientists to securely deploy custom web apps in their organisations
• improving data literacy in the NHS and helping stakeholders understand the amazing potential of R, Python, advanced analytics, operational research and data science
• how to teach coding, data science and operational research effectively

Here are a couple of the things I've created:

This video intermittently fails to load - try refreshing the page if it's showing an error!

Play around with a live demo of this here - note that it won't load in Firefox!

This custom PowerBI visual takes a simple, routinely collected dataset of patient history and displays it in a clear format.

Long-term referrals as well as individual interactions with services can be shown, with customisable tooltips, allowing clinicians to get an overview of a patient's history in a quick glance and spot repeated patterns of service interaction. Designed initially for use in mental health services, where clients can have more than a decade of data, thousands of contacts, and systems which display these details as rows of text rather than in a way that allows easy interpretation.

This custom PowerBI visual allows SPC charts to be created within PowerBI that automatically adhere to the making data count guidance.

Extensive customisation is available through built-in menus. Multiple output types are possible, including faceted graphs, metric cards, and several summary tables.