Gen AI Data Scientist

Technical Skills:

Programming Languages & Libraries: Python, OpenCV, Pillow, Numpy, Pandas, Scikit-learn, TensorFlow, Keras, PyTorch, Hugging Face, NLP, LLM, RAG, SVM, CNN, OpenAI, agentic ai.

LLM & GenAI Frameworks: Retrieval-Augmented Generation (RAG), LangChain, LangGraph, Transformers, TensorFlow, Keras, PyTorch, FastAPI, Hugging Face, crewai, langgraph YOLOv8 , Streamlit.

Cloud & Infra: AWS (S3, Lambda, Bedrock, OpenSearch, API Gateway, IAM), Infra automation, Docker

Vector Tech: Semantic Search, Context Similarity Search, Vector Databases (Pinecone, OpenSearch, FAISS) Programming & Integration: Python scripting, API development, Claude & Bedrock LLM integration

Databases: PostgreSQL, MySQL, Dynamodb

Qualifications: Bachelors in Technology / Masters in Data Science/ Masters in Mathematics/ Masters in Physics/ Statistics or any other quantitative field

Certifications: GCP/ AWS Certified Machine Learning/ Gen AI Engineer/ Modeller

Relevant Work Experience: 5+ years (modelling + Gen AI + Data Engineering)

Projects:

1. Gen AI

a. Led the design and delivery of a RAG solution to improve customer support efficiency by enabling rapid, accurate responses to policy, product and service queries from a centralized knowledge base. Led the end-to-end pipeline covering web ingestion, semantic embeddings, and vector retrieval (Pinecone), and integrated a GPT-4-class language model to power a context-aware AI assistant

b. Built enterprise-grade Q&A platform using RAG pipelines with AWS Bedrock, OpenSearch, Lambda, LangChain, LangGraph. Integrated Claude LLM for high-quality contextual responses with semantic search. Designed modular infra-ready architecture for scalable GenAI deployments. Worked with security teams to align AI solutions with compliance standards. Automated deployment workflows using Python and AWS services.

c. GenAI Q&A Platform (RAG + AWS + Claude) Built an enterprise chatbot using LangChain, LangGraph, Claude, and AWS Bedrock. Used semantic search with vector DBs to improve response precision and latency.

d. Built a conversational travel planner using bedrock claude sonnet and CrewAI, integrating MCP tools for itinerary and event suggestions across Indian cities. Demonstrated feasibility of tool-based agent collaboration and LLM intent handling, influencing product roadmap adoption. Automate customer travel inquiries, which led to a 30% increase in support efficiency and reducing the need for manual intervention

e. Designed and deployment of a robust Kubeflow MLOps framework on GCP Vertex AI, scaling big data streaming workloads in 5-minute intervals with automated feature pipelines, model training, and low-latency scoring. Built a centralised real-time Grafana dashboard along with Python-based automation to instantly onboard new network elements or KPIs, resulting in reducing dashboard development efforts by >90%. Delivered GenAI & RAG training sessions and presented project outcomes in global town hall, driving capability uplift and increasing leadership visibility.