Projects
I have added some projects I have worked/am currently working on below. When I look back at old code/work I do cringe with horror, but it is also satisfying having a record so I haven’t deleted or altered anything I’ve worked on. Take a look and get in touch with any questions.
Ecofit Homes
Python, FastAPI, MongoDB, Flask, Django
The main project I’ve been working on for the past year has been Ecofit Homes. We started as an end to end retrofit platform but quickly pivoted to being a data company as we realised the huge value opportunity here. I made a data cleaning pipeline, to automate the collection, cleaning and combining of our data sources. Then I made an API which delivered the data to clients from our MongoDB database, and finally a web platform to visualise results (a bit like PowerBI but much more custom functionality).
Academic Papers
I’ve been working as a co-author on a couple of academic papers, generally around the topics of AI/ML in healthcare. We have now published three papers together with more to come soon!
ChatGPT Versus Consultants: Blinded Evaluation on Answering Otorhinolaryngology Case–Based Questions
Chasing sleep physicians: ChatGPT-4o on the interpretation of polysomnographic results
A comprehensive review of using magnetic resonance imaging scans to detect the presence of MGMT methylation promoter in glioblastoma patients (2022)
Deep Learning, Tensorflow, Jupyter, Python, Streamlit, Pydicom
This was my masters dissertation project where I worked on the 2021 BRaTs kaggle competition to show why analysing biomarkers using mpMRIs is not possible. I did this by building a range of convolutional neural networks to assess their performance for identifying MGMT methylation promoter biomarker. I built on this by creating a proof of concept web app for clinicans to assess patient MRI scans.
Amazon Price Scraper
Python, Bash
I recently bought some new headphones full price and then Amazon discounted them a couple of days later 🥲. I have free returns for three months and want to check the Amazon price to see if it gets to the point where its worth me going to the store to return them and get them from Amazon. So I made a super simple scraper project running on crontab to scrape Amazon and get the current price of my headphones every hour.
Next I’ll get it to email me if the price dips below my “efficiency point” so I can return my ones and buy them from Amazon! The scripts are short and sweet but saves me some time and hopefully money!
Code available here.
HarryCoin - Making my own Cryptocurrency
Python, Flask, Postman
I’ve been intrigued by crypto for a wee while, but have never explored it particularly deeply. I own a tiny amount of bitcoin and have done for a couple of years but my understanding of how it works has been surface level. I haven’t written much Python in the last few months so I thought the best way to learn would be to build my own cryptocurrency from scratch in Python. My main guide is here, but I’ll be taking inspiration from other places I’ll reference on my github here. My stretch goal would be to host it on my Raspberry Pi but for now I’ll focus on writing the code.
Audio Event Detection for Emergency Service Vehicles (2020 - 2021)
Deep Learning, Tensorflow, Jupyter, Python
This was my final year dissertation project of my undergraduate computer science degree. I build 4 deep learning (DL) models and a web app to look at using audio for siren detection. The aim was to train models to deploy for a) accessibility and b) driverless cars, but was mainly tailored at educating me on DL. It was very fun getting a DL model actually working and classifying results and I learnt a lot on model types and DL development. I have included a link to my final report below.