Final-year BSc student at Anglia Ruskin University, Cambridge, working at the intersection of large language models and clinical safety.
Building oncology-llm-hallucination-benchmark — my dissertation project evaluating how frequently and how dangerously frontier AI chatbots (Claude, GPT-4, Gemini) hallucinate on cancer-related patient questions, benchmarked against NICE oncology guidelines. Execution May–August 2026.
Clinical NLP, LLM evaluation, patient-safety informatics, reproducible research.
Python · PubMedBERT · spaCy · scikit-learn · Anthropic, OpenAI, and Google AI SDKs
Email: bhaumikdg@gmail.com ORCID: 0009-0002-1736-3185 LinkedIn: www.linkedin.com/in/bhaumikdg