I build apps and games where AI does real work: chat assistants, smart search, content generation, personalisation and AI-driven gameplay. I connect Flutter and Unity front-ends to LLM APIs such as OpenAI and Google Gemini through a secure backend, and at Nplays I lead the architecture for AI-enabled gameplay products.
In-app assistants that answer questions, guide users and take actions, grounded in your own content and data.
Summaries, descriptions, quizzes and recommendations, plus search that understands what users mean.
Adaptive difficulty, AI opponents and sparring partners, and behaviour-tree enemies — from my DDA research testbed to VR combat training.
API keys kept server-side with Firebase and REST services, usage limits, moderation and cost control.
AI in my work
Projects where AI drives the experience — adaptive difficulty, AI sparring partners and behaviour-tree enemies.
All projects
Research simulation · Shooter
A third-person shooter testbed for dynamic difficulty research — bot profiles, rule-based and fuzzy DDA, and full metric export.
VR fitness · Combat training
A virtual combat gym: train on heavy bags, pads and AI sparring partners while every punch is measured for speed, power and accuracy.
Action RPG · Open world
An open-world action RPG with combo combat, behaviour-tree enemy AI, weapon upgrades and dynamic missions.
Tools & tech
FAQ
Chat assistants, smart search, summaries, content and quiz generation, personalised recommendations and AI-driven characters in games — built with LLM APIs such as OpenAI and Google Gemini.
Yes, if it’s built properly: API keys stay on the server, requests go through your backend with rate limits and moderation, and user data is only sent where it’s needed.
Yes — dynamic difficulty adjustment with rule-based and fuzzy logic, AI sparring partners in VR combat training and behaviour-tree enemy AI in an open-world RPG.
More expertise