About the Role

Base model selection and fine-tuning quality are the foundation everything else is built on — the default recipes here shape how well every downstream model performs. As a Fine-Tuning Research Scientist, you'll own that foundation.

What You'll Do

- Select and evaluate base models for fine-tuning suitability
- Design and maintain instruction-tuning and supervised fine-tuning recipes, including LoRA/QLoRA adapters
- Support customer-specific tuning requests as reference implementations for the wider platform
- Run and interpret fine-tuning ablations to keep our default recipes current with the field
- Collaborate with our evaluation team to validate fine-tuning quality against real benchmarks

What You'll Bring





- 5+ years of relevant experience in applied machine learning or NLP
- Robust hands-on experience fine-tuning large language models, including LoRA/QLoRA
- Proficiency with PyTorch and the Hugging Face ecosystem
- Comfort running and interpreting training ablations at scale
- Solid grounding in evaluation methodology, to know whether a fine-tune actually improved the model

Education

Master's or PhD in Machine Learning, Computer Science, Statistics, or a related quantitative field.

Nice to Have

- A master's or PhD in Machine Learning or NLP
- Experience fine-tuning models at an applied AI company

📌 Fine-Tuning Research Scientist (Bengaluru)
🏢 HCLTech
📍 Bengaluru