Srikar Kashyap Pulipaka

I’m an NLP researcher interested in multilingual NLP, synthetic data augmentation, and LLM evaluation. I build systems that make language models work reliably across languages and domains.

At ACL 2026, I’m presenting a poster on multilingual polarization detection (2nd place, SemEval Task 9) and giving a talk on toxicity detection in gaming chat (4th/35, EEUCA 2026).

I currently work at Cummins Inc where I build RAG and agentic systems for enterprise knowledge retrieval. Previously, I was a research assistant at the Kelley Data Science and AI Lab at Indiana University, working on LLM fine-tuning for mental health QA and stance detection. I hold an MS in Computer Science from Indiana University.

Research interests: Multilingual text classification, synthetic data generation and quality filtering, preference optimization, low-resource NLP, retrieval-augmented generation, LLM-as-judge evaluation.

Feel free to reach out — srikar.kashyap@gmail.com.