arindam vashishth

At Wells Fargo, I build distributed platforms for real-time risk calculation, market-data pipelines, and reporting systems. I’ve led migrations to Kafka-based event streaming that moved 80M+ records daily, tripled throughput, and reduced replication failures by 90%.

I’ve also taken agentic AI systems from prototype to production, including a RAG support assistant adopted by 30+ development teams and a multi-agent workflow that analyzes 1,200+ Jira tickets monthly across 20+ repositories.

I design systems around concurrency, reliability, and operational efficiency. My work includes a Java and Spring Boot scheduling framework supporting 1,000+ daily jobs, Splunk dashboards across 100+ servers, and Apache Ignite improvements that reduced weekly downtime from five hours to 30 minutes.

Earlier at Siemens, I built graph neural network models for power-grid load-flow prediction using PyTorch Geometric. I also explore machine learning and reinforcement learning for stock prediction and portfolio analysis.