Newspaper Chatbot (Bellingchat)
Uses newspaper articles and an LLM give a human-readable response. .
Here's the Medium article
I'm a Data Scientist
I have experience in: AWS, NodeJS, LLMs, Pandas, Python, SQL, TypeScript
Uses newspaper articles and an LLM give a human-readable response. .
Here's the Medium article
Online dashboard for Bellingcat showing Instagram posts in given GPS coordinates
Predict the sentiment of a movie review using NLP. Written using FastAI library.
Here's the Medium article
Type your favourite movie and find more like it. Uses collaborative filtering
Here's the Medium article
Upload a picture of someone to determine whether or not they're wearing a mask. Model trained in FastAI.
Here's the Medium article
I wanted to find more about my favourite Spotify aritsts. I used the Spotify Web API to them and then the MusicBrainz API to find artist metadata like age, country of origin and genre.
As part of a Kaggle competition, used sci-kit learn ML library to predict the outcome of a passenger based on their attributes.
NeoDyne's company website. Developed in WordPress, much of the work done with the Managing director of the company.
I made a device to monitor the wellbeing of an infant remotely, by reading the direction and acceleration using a BBC Microbit. This was then sent over Bluetooth Low Energy (BLE) to a Raspberry Pi. Using Noble, a NodeJS library, data was stored in an Influx database and published to an MQTT topic. A frontend webpage then subscribed to the topic.
I designed and implemented a system to record the energy used in a home over the course of a day. The trends were then be used to maximise the usage of renewable energy sources like solar. Energy usage was collected with smart plugs and analysed with Python, then stored in an Influx database.