Pavan Kumar Malasani

About

GenAI/ML engineer in Citigroup’s finance environment who has deployed production RAG systems for investment banking under strict privacy and model-risk constraints. Built an internal-VPC Llama2 + Pinecone + LangChain solution with NER redaction and citation-based verification to prevent hallucinations, delivering major time savings, and also partnered with global finance executives to ship an AI early-warning indicator for treasury/liquidity risk.

Experience

Education

Key Strengths

• Built and deployed secure, internal-VPC GenAI RAG system on Kubernetes for investment banking market briefs
• Implemented privacy-preserving pipeline with NER-based sensitive data redaction (no public API exposure)
• Designed LangChain verification layer enforcing per-claim citations to retrieved sources to reduce hallucinations
• Established evaluation framework using faithfulness/answer relevance, human-in-the-loop audits, and automated evaluator models
• Strong regulated-finance governance: audit trails, traceability, bias testing, and MRM-aligned AI lifecycle processes
• End-to-end orchestration across infra (Kubernetes), data workflows (Airflow), and agent logic (LangChain)
• Production monitoring and reliability practices: MLflow drift monitoring and automated retraining triggers
• Delivered measurable impact: automated ~28 hours/week of analyst research; cited 40% reporting time reduction
• Effective translation of model outputs into executive-ready metrics and dashboards for treasury risk stakeholders