PROJECTS
This is an ongoing blog on artificial intelligence that I write. The goal is to demystify AI as much as possible, report of some interesting developments as well as give you ideas of how you can implement AI effectively in your business (it's a lot easier to get started than most people think!).
.png)
MAIT stands for McCabe's Artificially Intelligent Tipper, and it's an AI-driven system for predicting the winner of sporting contests. Appearing in the media in over 75 countries in no less than fifteen languages, this has been a super popular project over the years.
This project uses techniques from natural language processing to assess the similarity of documents (particularly applicant resumes vs job ads) in a meaningful way, even if they potentially don't have any words in common. It's a great way to assess the best candidates for a particular role, or for a candidate to see which jobs they are best suited for. Contact me for more details.
I've been using GPT-3 (now GPT-4) and ChatGPT for some time, getting some amazing results. Most specifically in the mental health space, and this is something I'm hoping to announce soon. I've also used language models to create tool called Ghost Writer for writing automated blog articles.
This was the subject of my PhD thesis and it was a great experience to be able to deep dive into all elements of neural networks/deep learning, hidden Markov models and biometrics. Spending years understanding how everything works under the hood I was able to build some very accurate biometric tools that are still in use. Take a look at some of my academic papers on the "Published" Page or if you're interested in my book check it out on Amazon or contact me directly.
There are some great implications of being able to accurately predict peaks or troughs in a noisy numbers series (stock market prices for example!). This innovative approach is one of the algorithms I've developed, and it's been successfully implemented in such diverse areas as neural network training optimisation, detection of acceleration turning points and detecting unusual physiology events in SIDS early warning research.