Data Scientist - Generative AI (LLM | RAG | Agentic AI)

The purpose of this role is to design, build and deploy Generative AI solutions that solve real business problems across TVS Credit. From building LLM-powered chatbots, RAG pipelines and multi-agent (agentic) workflows to fine-tuning language models and integrating them into production systems, the role focuses on turning cutting edge AI capabilities into reliable, scalable products that improve customer experience, strengthen fraud detection and drive operational efficiency.

Key Responsibilities:

• Design, build and deploy Generative AI solutions LLM/SLM-powered chatbots, RAG pipelines and multi-agent (agentic) workflows for enterprise use cases
• Develop and integrate LLM applications using frameworks such as LangChain / LangGraph, including prompt engineering, tool / function calling and orchestration of multiple agents
• Build retrieval-augmented generation (RAG) systems generate embeddings, manage vector databases (e.g. pgvector, FAISS) and implement hybrid (semantic + keyword) search over structured and unstructured data
• Fine-tune and evaluate open-source language models (SLMs / LLMs); optimise inference, latency and cost; and implement guardrails to control hallucination and ensure safe, grounded outputs
• Develop and expose AI services as production-grade REST APIs (e.g. FastAPI), integrate with enterprise databases (Oracle / PostgreSQL) and object storage, and support deployment on GPU / cloud infrastructure
• Work with OCR and document-intelligence pipelines to extract, validate and reconcile information from identity and business documents
• Develop proof-of-concepts to validate new AI approaches, interact with vendors / partners on scope, and productionise successful POCs

Qualifications

BE / BTech / MTech / MSc / MCA in Computer Science, Data Science, Artificial Intelligence / Machine Learning, Statistics or any related quantitative field. 12 years of experience building AI / ML or Generative AI applications (strong project portfolios from fresh candidates may also be considered).

Technical Skills (Must Have)

• Strong programming skills in Python
• Hands-on experience with LLMs / Generative AI and prompt engineering
• Agentic AI frameworks LangChain, LangGraph, OpenAI / LLM APIs
• Embeddings and vector databases (pgvector, FAISS or similar) and RAG design
• Databases and SQL PostgreSQL, MySQL, Oracle DB
• Building / consuming REST APIs (e.g. FastAPI) and version control with Git

Preferred / Bonus Skills

• Convolutional Neural Networks (CNNs) and computer vision face detection / recognition, image-quality assessment and OCR
• Fine-tuning of SLMs / LLMs and model evaluation