At BlueOptima, I build production Generative AI systems that generate more maintainable code using open-source and closed-source LLMs, including CodeGemma, Llama 3.x, Mistral, GPT-4, Gemini, and Claude.
I shipped a model-agnostic code maintainability AI agent with automated validation and quality checks, using tool calling to make context-aware code-generation decisions without manual effort.
I've also built LLM benchmarking pipelines across providers for enterprise coding tasks, RAG systems for clinical PDFs, and ML pipelines that identify drivers of developer productivity.
My work spans Python microservices, developer-behavior clustering, predictive modelling, dependency security recommendations, and Grafana dashboards that turn engineering data into actionable insights.
