Sanchit Kumar Mishra

I'm contributing to enterprise-scale AI systems at Ethara AI, where I build domain-specific technical datasets for LLM fine-tuning across reasoning, coding, and instruction-following tasks.

I design evaluation pipelines for factual accuracy, coherence, and alignment, and run prompt engineering experiments across instruction styles and edge cases. I also work with cross-functional teams to improve annotation efficiency, dataset quality, and model reliability.

I've built healthcare analytics dashboards processing 55,000+ records and reducing manual reporting effort by nearly 65%, and developed an NLP extractive summarization system that reduced document review effort by approximately 60–70%. My published research explores scalable healthcare analytics using Python and Power BI.