• Owned the iOS execution pipeline end-to-end (vision reasoning, screenshot capture, action planning, on-device execution), enabling production-grade iOS test automation alongside Android. • Improved the existing execution architecture to scale reliably to ~300–500 executions/day with 50–60 concurrent long-running test plans (3–4 hours each). • Built and owned the App Management module from scratch, enabling self-serve onboarding for apps, devices, credentials, and execution settings; used by 30–40 customer teams. • Integrated and stabilized execution across third-party device providers (Appetize for iOS, LambdaTest for real devices), resolving API mismatches, emulator instability, and device-state failures through retries and safeguards. • Improved execution reliability and performance, implementing failure-handling for device and provider errors and reducing median vision-execution step latency from ~15s to ~5s, achieving ~90% end-to-end test plan success rate. • Business impact: iOS execution support enabled US-focused pilot availability, leading to ~5 new US customer teams per week and rapid platform adoption.

• Owned and built the inventory management backend from scratch for a mission-critical fresh-produce supply chain, supporting 500+ IoT-enabled terminals operating concurrently during synchronized procurement and distribution windows. • Designed and implemented the low-level system architecture using Java (Spring Boot) and MySQL, independently driving data modeling, transaction boundaries, and service interfaces, with final design reviews at VP level. • Engineered a 24/7 production system where inventory accuracy directly impacted real-time procurement and dispatch; planned deployments and maintenance during non-operational windows to sustain ~95–98% uptime during peak business hours. • Built real-time inventory workflows to handle frequent quantity changes caused by spoilage, quality variation, and operational adjustments, ensuring correctness under concurrent updates across all active terminals. • Designed and developed the Rate Card Management system from scratch, enabling dynamic pricing across time windows, geographies, and product categories; actively used by 5,000+ internal users across operations, business, and data teams. • Worked closely with product, operations, and on-ground field teams, including site visits across Tamil Nadu, to translate real operational constraints into scalable backend workflows and automated reconciliation processes, saving 6+ hours per day for operations teams.

Optimized cloud security platform by implementing Redis caching, reducing query latency by 20% and improving system performance.

Enhanced Couchbase query efficiency, cutting query time by 50%, ensuring faster and more reliable threat detection.

Secured first prize as team leader at the Product Expo.

Enhanced expertise in data engineering and backend development.

Collaborated closely with leadership and PMs to fine-tune project requirements.