May 2022 – Apr 2025 · 2 yrs 11 mos · Bangalore Urban, Karnataka, India

• Tech Stack: Java, Spring Boot, Kafka, Spark, AWS (SQS, Lambda), MongoDB, Redis, MySQL, Docker, Jenkins, TDD, BDD, Python, Snowflake (Beginner)
• Designed and deployed scalable Spark jobs to sync 20M+ daily records between Marketo and Dynamic Chat, tuning Kafka backpressure to prevent message loss and reduce job failure rate by 60%.
• Reduced DB latency by 35% by implementing connection pooling and concurrency-safe patterns, improving throughput for concurrent user sessions by 2.5x.
• Improved agent permission workflows, achieving 95%+ success rate in sync jobs using parallel processing and DLQ-based fault isolation—reduced permission update failures by 80%.
• Spearheaded WhatsApp integration from PRD to deployment, resulting in a 30% increase in user engagement via chat, with 100% SLA adherence in message delivery and opt-in tracking.
• Contributed to Project Hummingbird, launching GenAI chat capabilities that handled 40K+ conversations/month, with a 24% deflection rate, reducing load on live agents.
• Led development of LiveChat platform supporting:
• Real-time chat latency <300ms
• 80%+ successful playbook-triggered meetings booked
• Seamless agent transition based on NLU score thresholds
• Reduced campaign deployment time by 70% through a centralized ad network configuration system that allowed runtime updates without redeployment.
• Built an AWS event-driven backend (SQS + Lambda) that handled 100K+ events/day, with 99.99% uptime, and scaled to meet burst loads of 5x without manual intervention.
• Developed secure Auth microservice supporting 1M+ sessions/month with IMS token verification and fallback support—achieved 99.95% token validation success rate.
• Resolved 20+ P1/P2 tickets on first contact, improving team’s ticket resolution rate to 98% and reducing average response time from 3 hours to 45 minutes.

Jun 2021 – May 2022 · 11 mos · Gurugram, Haryana, India

• Designed and developed the Master Price API, handling 10,000+ QPS, improving real-time price fetching and enrichment for 50+ financial institutions, reducing latency by 40% in pricing updates.
• Redesigned the Pricing Engine to compute EOD prices and PnL using Black-Scholes and yield-curve models, increasing computational accuracy by 25% and reducing processing time by 30%.
• Migrated ActiveMQ to a hybrid AMQ/SQS model, improving scalability and reducing operational overhead by 35%, while enhancing availability and uptime for op/cloud-native clients.
• Developed APIs for price enrichment, integrating multiple pricing models using the Adapter Design Pattern, enhancing system adaptability.
• Automated end-of-day pricing finalization with scheduled batch processing, reducing manual intervention by 80% and ensuring on-time report generation for 100+ users in real-time trading.
• Implemented fault-tolerance techniques like circuit breakers, retry mechanisms, and DLQs, improving system resilience and reducing downtime by 99%, ensuring consistent availability during peak trading hours.
• Optimized system performance by enhancing SQL queries, implementing caching, and improving multi-threading, reducing API response time by 30% and increasing throughput to handle 1,000+ concurrent requests.
• Coordinated with quant and trading teams to align on PnL calculations, validate assumptions, and ensure the pricing models met business requirements, supporting $100M+ in daily transactions.
• Developed a real-time PnL computation platform, aggregating trade data and market snapshots, enabling real-time decision-making for trading teams, improving trading efficiency by 20%.
• Implemented test coverage using JUnit and Mockito, integrated into CI/CD pipeline, reducing production defects by 40% and improving release stability.
• Validated pricing models and PnL results through market data comparisons and stakeholder feedback, ensuring 99.9% accuracy in final calculations for financial reporting.

Jan 2021 – Jun 2021 · 5 mos · Bengaluru, Karnataka, India

• Worked on the RLC (Radio Link Control) layer of Samsung’s 4G eNB system for Reliance Jio. Involved in debugging, feature enhancement, and field issue support based on the 3GPP LTE specification.
• Key Contributions:
• Implemented feature enhancements for RLC throughput logging and bearer-level statistics.
• Resolved critical field issues like memory leaks and incorrect KPI reporting.
• Supported SIT/SVT teams in debugging and analyzing protocol-related issues.
• Helped in initial lab setup and created basic scripts for log collection and automation.

Aug 2020 – Dec 2020 · 4 mos · Gurugram, Haryana, India

• Integrated Real-Time Video & Interactive Tools: Seamlessly integrated GeoGebra and 100ms for live interactive classes, enabling real-time video calls and interactive whiteboards for over 200,000 users.
• Built Scalable Service Architecture: Developed production-ready schedulers, event processing, and batch processing systems, handling over 100,000 events daily, ensuring high availability and performance for critical business services.
• Developed Monitoring Systems: Created real-time health monitoring systems with automated alerts, reducing system downtime by 30% and improving overall reliability.
• Database & Backend Services: Utilized Firebase Firestore to manage dynamic meeting rooms and role-based access (teacher/student), optimizing performance and ensuring smooth learning sessions for users.
• Optimized Cloud Infrastructure: Leveraged AWS services (Aurora, EC2, S3) to support a growing user base, ensuring scalable data handling and efficient backend services.
• Backend: Django, Firebase Firestore, AWS EC2, Aurora, S3, cloudwatch
• Frontend: Vue.js
• Real-Time Services: 100ms, GeoGebra

Jan 2019 – Jan 2020 · 1 yr · Jaipur, Rajasthan, India

• Research on android sensors. Applied deep learning models to Crack phone password.