Hariteja Vemu

Built a natural language-to-SQL agent using Claude 3.5 Sonnet and LangGraph optimizing customer communications Developed an MCP server with Redis-based semantic search, data analysis, web search, image search tools Engineered a risk categorization model using XGBoost and GridSearchCV, to score merchant fraud risk Implemented full inference profiling with Langfuse across AI pipelines, tracking computational cost and latency Developed a RAG based chatbot using Redis Vector Database and HuggingFace Embeddings

Data Science : Designed a revenue prediction model using XGBoost Automated sentiment-to-rating mapping for customer feedback using a fine-tuned RoBERTa model Built an interactive sales assistant with RAG and FAISS Developed an automated KYC verification pipeline with OCR and OpenAI Data Engineering : Engineered scheduled ETL pipelines with Airflow, integrating data from MongoDB, MySQL, and S3 into centralized warehouses to support real-time dashboards and business intelligence.