.png){kind=logo}
nitesh bommisetty
Mid-level Data Scientist specializing in ML, NLP, and LLM-powered solutions
Connect with nitesh
nitesh already has a relationship with Reval, so a warm intro from us gets a much better response than cold outreach.
Recommended
Already have an account?
About
AI/NLP-focused practitioner who built a zero-/few-shot LLM event extraction system on the long-tail Maven dataset, combining prompt-structured outputs with LoRA/QLoRA fine-tuning and rigorous F1 evaluation. Also implemented entity resolution/data cleaning pipelines and embedding-based semantic search using Sentence-BERT + FAISS, and has healthcare experience delivering a multilingual speech/translation mobile prototype using HIPAA-compliant Azure Cognitive Services.
Experience
Education
Key Strengths
- Built zero-shot/few-shot LLM event extraction framework on Maven (168 event types, long-tail imbalance)
- Improved rare event-type performance via sampling strategies and increasing few-shot examples
- Fine-tuned open-source LLMs using LoRA/QLoRA with Hugging Face Transformers
- Designed multi-step data cleaning + entity resolution pipeline (token index alignment, deduplication, NER-based merging)
- Implemented embedding-based semantic linking/search using Sentence-BERT + FAISS with precision/MRR evaluation
- Production-minded Python data workflow design (modular stages, schema checks, chunking/parallelization, PySpark scaling)
- Pragmatic model selection for healthcare app delivery (chose Azure Cognitive Services over self-hosted Whisper due to reliability/compliance/timeline)
Discover more candidates like nitesh
Search across thousands of pre-screened, high-quality, high-intent candidates on Reval.
Search Talent