At Inforens, I engineered a production RAG-based document intelligence system that retrieves information about universities, tuition fees, scholarships, and admission eligibility. I connected semantic search and LLM-powered question answering to the platform through Python and FastAPI endpoints supporting the Ask Nori chatbot.
I also built automated ingestion pipelines for university APIs, web scraping, and PDFs, with incremental embedding updates to avoid unnecessary processing and full-index rebuilds. I configured AWS infrastructure, containerised applications with Docker, and monitored logs and performance with CloudWatch.
To evaluate response quality, I implemented automated RAG assessments using Ragas, faithfulness metrics, and test datasets. Earlier, at Motleyscape, I developed ETL pipelines and Power BI dashboards for real estate and construction reporting; my MSc dissertation project was a service provider recommendation engine.
