Selected intern's day-to-day responsibilities include:

- Develop LLM applications using Retrieval-Augmented Generation (RAG) pipelines for accurate, context-aware outputs.
- Fine-tune open-source LLMs using PEFT/LoRA/QLoRA. Perform post-training using different methods, including RL and DPO.
- Validate and test custom LLM-based applications and/or fine-tuned models.
- Design and optimize semantic search systems using embeddings, vector databases, and ranking models.
- Analyze datasets, identify gaps in quality and coverage, and improve them systematically.
- Design experiments (A/B testing, offline evaluation) to validate system performance.
- Collaborate with AI engineers and domain experts to integrate systems into real-world applications.




- Create and maintain technical documentation.

Who you are
- Someone with a solid technical background and an interest in LLMs and machine learning, specifically NLP.
- Proficiency in Python (experience with ML frameworks is a plus).
- Understanding of LLMs, embeddings, vector search, RAG architectures, and different post-training methods.
- Familiarity with search systems and ranking algorithms.
- Knowledge of Deep Reinforcement Learning, DPO, RLHF and RLVR would be plus
- Having higher qualifications, such as an MTech or a higher degree, is a plus.

📌 GenAI Intern (Bangalore Urban)
🏢 Nosh Robotics
📍 Bangalore Urban