👋 Hey, I'm Dweep Pandya
Senior AI Engineer

📍 Pune, Maharashtra

• Led end-to-end architecture of an Enterprise-Scale RAG system on Azure Databricks — designed for high scalability, security, and low-latency performance.
• Orchestrated complex data pipelines for unstructured data processing, implementing sophisticated chunking and vectorization strategies.
• Engineered context-aware AI capabilities exposed as Model Context Protocol (MCP) endpoints for seamless agentic interoperability.
• Designed hybrid search pipelines combining dense vector and sparse retrieval for high-fidelity document retrieval.
• Implemented dynamic context assembly and advanced prompt optimization techniques to maximise LLM output quality.

📍 Pune, Maharashtra

• Led a team of 4 Data Scientists, driving technical mentoring, sprint planning, and cross-functional collaboration.
• Delivered $1M+ USD in ROI for clients using AI-driven Operational Excellence and Sustainability solutions.
• Spearheaded company's transition to Generative AI — built RAG-based agentic architectures using LangChain, Vector DBs, and fine-tuned open-source LLMs.
• Architected scalable production ML infrastructure using Terraform (IaC), Kubernetes (Kubeflow), and MLFlow.
• Built predictive maintenance, energy forecasting, and anomaly detection models at scale on AWS (EC2, ECS, S3, EKS, Lambda, Glue).

📍 Pune, Maharashtra

• Developed and deployed supervised/unsupervised ML models for Fraud Detection and Anti-Money Laundering (AML) on AWS SageMaker.
• Reduced manual analytical effort from 6 weeks to 2 hours through ML-based process optimization — massive ROI impact.
• Managed major client accounts directly, including Goldman Sachs.
• Worked extensively with big datasets (100M+ rows) using Athena, S3, and Python-based pipelines.