Jun 2023 – Present · 2 yrs 9 mos · Bengaluru, Karnataka, India · On-site

• Led Oracle TxEventQ’s AI enablement by integrating in-database ML inference and vector similarity search for
• RAG-based AI agents. Deployed LangChain NLP workflows in MCP Server, improving inference throughput by 40%
• and reducing LLM calls by 30%, enabling real-time diagnosability and AI-driven workload optimization.
• Designed a Complex Event Processing (CEP) engine delivering sub-50ms correlation latency across 500K+
• events/sec, using NFA-based pattern recognition, state backends, and checkpointing. Implemented memory-efficient
• state pruning and reclamation policies for long-running sessions.
• Re-architected the Rule Engine to support 10M+ TPS workloads, achieving 40% memory optimization via global
• caching and data-locality techniques. Built a lightweight wrapper increasing internal adoption by 3×, and resolved
• mutex contention across 60K+ concurrent sessions using adaptive object replication in C.
• Developed a high-throughput stream processing engine integrating Flink/Kafka connectors for distributed event
• streaming, achieving 12M events/sec at ¡10ms latency using Java concurrency and custom memory pooling.
• Showcased an Agentic AI demo at Oracle AI World 2025, featuring a LangChain-integrated event pre-processing
• framework that reduced LLM hallucinations by 35% and query cost by 25%, positioning TxEventQ as a cognitive layer
• for AI workflows.