- Developed prompt injection defence models using machine learning to safeguard LLM-integrated applications- Created large, high-quality custom datasets by prompt engineering large language models to synthetically generate data

- Developed multiple AWS automation scripts which were responsible for manipulating and interacting with hundreds of AWS instances - Developed RESTful APIs for an internal company application, facilitating the allocation of employee projects. Integrated robust user authentication systems, alongside streamlined data submission and retrieval functionalities- Facilitated the seamless transition of the company's new support role as data engineers and analytics for the Home Office by conducting comprehensive playback sessions and providing support throughout the transition phase

- Worked on the development of a demand forecasting model for inventory management- Performed data exploration, cleaning and analysis of the historical raw dataset to turn it into useful data- Trained a LightGBM regression model that achieves an error rate of less than 1% - Deployed the ML model into a fully automated machine-learning pipeline with Azure

- Fixed issues that were reported on the various sites we monitored. These varied from hardware-related problems on the AV rack and smart devices, to software-related issues, where building control systems needed to be reprogrammed or configured- Installed and programmed 100s of IoT (Internet of Things) devices in commercial and residential buildings using programming languages and software such as C#, Simpl, Crestron, Dallmeier- Proposed bespoke designs which integrated smart IoT devices for residential and commercial buildings. Adaptability was vital in order to meet client demands and supply limitations, ensuring cost and time efficiency