Feb 2023 – Present · 3 yrs 1 mo

• Building an AI-powered performance framework using LangGraph and AI agents to automate log analysis, detect regressions, and cut manual triaging time by 70%.
• Reduced homepage API latency by 60% by implementing a scheduler-driven refresh-ahead caching strategy with Redis (ElastiCache) to proactively warm the cache.
• Developed machine learning models (XGBoost, Facebook Prophet, CNN) to predict sales trends and optimize inventory planning across millions of transactions.
• Created an AI agent using LangGraph, LangChain, and Groq APIs, backed by a lightweight SQLite memory store, to power intelligent workflows during a hackathon project.
• Modernized QBO’s inventory system as a key contributor by migrating it to a high-throughput service (432M+ daily transactions), enabling new features and generating $100K+ in add-on revenue.
• Reduced database footprint by 30% and increased ingestion throughput by optimizing data models and query patterns, leveraging polyglot persistence.
• Applied Domain-Driven Design and Onion Architecture to ensure modular, maintainable systems.
• Implemented a flexible channel disconnection feature using Chain of Responsibility and pessimistic locking, enhancing control across e-commerce integrations.

Feb 2021 – Feb 2023 · 2 yrs

• Led asynchronous communication via Apache Kafka, managing 5000TPS, ensuring data integrity in high-throughput scenarios using outbox patterns, Dead Letter Queues (DLQ), and idempotency.
• Orchestrated the migration of 600GB of real-time data using Apache Spark from a data lake to RDS, resulting in a 20% faster data processing time and improved data accessibility.