PAVANI SEELAM

Architected and developed AI-driven platform services for evaluating multi-step workflows, simulating real-world automation and decision systems. Built agentic execution pipelines integrating LLMs, APIs, and contextual memory for dynamic workflow orchestration. Designed scalable FastAPI/Node.js microservices handling 2M+ transactions/month with 99.9% uptime on AWS. Developed React/TypeScript dashboards to visualize system behavior, evaluation outputs, and performance metrics. Implemented secure authentication and authorization systems (OAuth2, JWT, RBAC) for multi-tenant workflows. Integrated LLM-based reasoning systems into production workflows, improving automation accuracy and decision-making efficiency. Built CI/CD pipelines (GitHub Actions) enabling rapid iteration and reducing regression defects by 42%. Implemented observability (CloudWatch, logging, tracing) reducing incident detection time by 60%. Optimized distributed workflows, reducing execution latency by 38% under high con

Designed and developed scalable microservices architecture supporting 1M+ transactions/day for financial systems. Built full stack applications (Angular + backend APIs) for workflow management and operational monitoring. Led migration from monolith to microservices, improving system scalability by 50%. Implemented CI/CD pipelines (Jenkins, Docker) reducing deployment time by 40%. Designed distributed workflows ensuring fault tolerance and high availability under peak loads. Built ML inference APIs (FastAPI + Azure AI) enabling intelligent decision-making systems. Optimized backend performance using Redis caching and query tuning, reducing response time by 30%. Resolved race conditions and concurrency issues, implementing idempotent APIs and transaction safety mechanisms. Participated in on-call rotations, troubleshooting production outages and improving system reliability through RCA. Collaborated with cross-functional teams to define system architecture, API contracts, and long-term s