Aug 2022 – Present · 3 yrs 7 mos · Hyderabad, Telangana, India

• Software Engineer, Agentic AI Infrastructure
• Led the design and development of core infrastructure for my org's Agentic AI systems. My work focuses on building the scalable, reliable, and observable platforms that enable agents to reason, remember, and act in complex environments.
• Architected and led the development of the core Session Service and Agent Orchestration Engine, providing the stateful, persistent backbone for complex, multi-step agent execution and reasoning loops.
• Designed and scaled the high-throughput Memory Service, leveraging vector databases and RAG techniques to equip agents with long-term memory and dynamic contextual understanding.
• Spearheaded a unified Prompt Management & Optimization platform, combining a version-controlled Prompt Externalization service with an automated pipeline to systematically test, tune, and improve agent performance.
• Played a pivotal role in building the Agentic Evals framework, creating robust evaluation systems, metrics, and data pipelines to measure task completion, tool usage, and AI safety.
• Engineered the AI Observability stack for agentic systems, implementing distributed tracing and specialized logging to effectively debug and monitor non-deterministic, large-scale agent behavior in production.
• Software Engineer, Machine Learning
• Transitioned from general software engineering to classical machine learning, focusing on building and deploying production-scale models.
• Developed and maintained end-to-end MLOps pipelines for model training, evaluation, and serving, handling terabyte-scale datasets.
• Engineered core components for large-scale feature stores, improving feature computation, consistency, and availability for multiple product teams.
• Collaborated with research scientists to productionize models for classification and forecasting, leading to a 97% improvement in Onboarding Experience for Employees, XWS