Hire Surendra - Vector Database Engineer

Architected and deployed 5+ production-grade GenAI solutions including RAG-based document intelligence and autonomous agents, reducing manual data processing overhead by ~50% across department workflows. Engineered advanced RAG pipelines utilizing LangChain and Vector Databases (Pinecone/Milvus), implementing MMR and Cross-Encoder Re-ranking to improve retrieval precision by ~35% and resolve multi-document context fragmentation. Developed agentic workflows using LangGraph and CrewAI, enabling complex task decomposition and self-healing logic that improved multi-step reasoning accuracy by ~40% compared to linear chains. Optimized LLM performance and cost-efficiency by implementing semantic caching and prompt compression, resulting in a ~30% reduction in API token expenditure and decreasing Time-to-First-Token (TTFT) by ~45%. Integrated Databricks-based data pipelines with GenAI architectures, leveraging prior expertise in Spark and Delta Lake to curate high-quality datasets for LLM fine-tuning and evaluation. Hardened AI application security by implementing robust guardrails (Nemo-Guardrails/Pydantic) and sandboxed code execution environments, ensuring 100% mitigation against prompt injection and unauthorized data access. Established LLMOps and evaluation frameworks using RAGAS and G-Eval, enabling automated testing of "Faithfulness" and "Relevance" which boosted deployment confidence and reduced hallucinations in production by ~25%. Built custom visualization agents that bridge the gap between Natural Language and Business Intelligence, automating the generation of interactive data plots and decreasing ad-hoc reporting time by ~60%.

* Provided technical consulting and custom software solutions for small-to-medium enterprises, focusing on cloud migrations and mobile optimization. * Co-managed a construction business venture, overseeing operations, resource planning, and digital transformation of internal tracking systems. * Focused on upskilling in Data Engineering (Spark/Azure) and early-stage LLM research, building foundational expertise in scalable data pipelines and NLP systems.