Worked on improving an AI chatbot platform by testing conversation flows and multi-turn dialogue handling. Used GPT-4 and LangChain to design prompts that could keep context from earlier messages. Helped set up evaluation tests to compare different approaches and identify the most accurate and natural responses.

Contributed to a system that turns text prompts into images and short videos. Integrated APIs such as Stability AI and Sora for generating high-quality outputs. Configured server-side processes to manage GPU workloads efficiently, ensuring fast response times even with high demand.

Built features to read AWS and Azure billing data, then explain it in clear, simple language. Created a natural-language query tool so users could ask specific cost-related questions. Added anomaly detection using basic machine learning models to identify unexpected spending patterns.

Helped develop a content generation platform producing text, images and audio. Built workflows to fine-tune AI models so outputs matched brand requirements. Assisted in connecting the backend to a web-based interface, allowing marketing teams to create and preview content in real time.

Worked on an AI writing assistant for e-commerce businesses. Implemented a semantic search function to quickly find similar content and added validation steps to check quality before publishing. Helped optimise backend processes so text generation was faster and more reliable.

Developed tools to process large health research datasets for longevity studies. Integrated GPT-based agents to summarise medical literature into concise reports. Built API endpoints to deliver results to an internal dashboard, making it easier for researchers to review and analyse data.