MuhammadHamza.

An autonomous agentic AI system for personalised learning. Students can learn anything in their own style, and the agent adapts to their cognitive ability — built with the OpenAI Agents SDK, RAG, and a vector store over reference material.

• →Autonomous agentic system that adapts to individual learning styles.
• →Personalised algorithms that adjust to learner cognitive ability.

Architecting customer & corporate SPAs for Mahaana — pension onboarding, portfolio dashboards, and corporate admin portals over ASP.NET Core microservices and Azure Service Bus.

• →+70% pension onboarding throughput for 15,000+ users.
• →−40% release cycle time via AI-assisted dev + Azure DevOps CI/CD hardening.

Designed an in-flight entertainment platform with content management and high-performance cloud infrastructure. Selected by PIA via public tendering for portable IFE devices on A-320 and B-777 aircraft.

• →−60% OneDrive→S3 transfer latency via async I/O + S3 multipart upload.
• →CMS for content providers to deploy on the DreamStream platform.

Architected an AI-powered platform that automates QA workflows using OpenAI's LLMs and a microservices architecture. Drove a 50% reduction in manual test-writing effort.

• →−50% manual test-writing effort via LLM-driven story-to-test generation.
• →Vector databases for efficient AI model integration.

Government platform for educational initiatives. Handled 500,000+ student registrations with Next.js 13, PostgreSQL, Redis caching, and a hardened Nginx + Docker deployment.

• →500,000+ student registrations.
• →Robust caching strategy with Redis for high-concurrency endpoints.

Healthcare administration system that uses AI to ensure fairness, transparency, and efficiency in healthcare access for financially disadvantaged patients. Recognised internationally at the Lokahi hackathon (lablab.ai).

• →AI-driven donation allocation prioritising urgent cases fairly.
• →International recognition from Lokahi (lablab.ai) technical mentors.