• Built and deployed churn prediction and CLV models using XGBoost, Ridge/Lasso, and Spark SQL, enabling proactive retention strategies that improved targeting accuracy by ~6% and increased upsell conversion opportunities by 4%.

• Designed and tested an AI-driven recommendation engine on AWS EMR using collaborative and hybrid ML techniques, improving click-through rates by over 7% while reducing low-relevance recommendations during pilot rollout.

• Implemented NLP-powered RAG pipelines with FAISS and LLMs to streamline data validation and reporting workflows, reducing manual effort and improving knowledge retrieval accuracy for Tableau-based executive insights.

• Scaled Oplay.ai, a multi-tenant AI automation platform, designing context-aware workflows and prompt orchestration to support diverse enterprise use cases across operations, analytics, and customer experience.

• Integrated and optimized multiple LLMs (OpenAI, Claude, Mixtral) using LangChain and RAG pipelines with Pinecone, reducing inference latency by ~35% while maintaining high response quality and consistency.

• Developed reusable prompt templates, evaluation frameworks, and guardrails to standardize model behavior, improving reliability and performance across multiple AI-driven applications.

• Engineered a high-performance backend using FastAPI, PostgreSQL, and Redis, handling 50K+ daily API requests with sub-200ms latency, while partnering with product and engineering teams to accelerate feature delivery.

• Maintained scalable ETL/data pipelines using SQL, Python (Pandas), and PySpark to process and standardize 120K+ daily payment and settlement transactions across multi-region systems, enabling reliable downstream reconciliation and reporting.

• Orchestrated AWS Glue ETL workflows integrating transactional, ledger, and reporting datasets, reducing month-end close cycles by 40% and improving financial reporting efficiency for finance stakeholders.

• Implemented data quality validation and anomaly detection frameworks using SQL to identify duplicate transactions, missing settlements, and revenue leakage, improving data accuracy by 30%+ before executive KPI reporting.

• Performed financial data analysis and forecasting using Advanced Excel (Pivot Tables, Power Query, XLOOKUP, scenario modeling), reducing forecast variance by 15% and enabling more accurate revenue planning and decision-making.