Software Engineer (B.Tech, Computer Science) with 2+ years of technical engineering experience building distributed systems and online services in Java, C, Python, and Go. Strong in relational databases (PostgreSQL, SQL), query performance and indexing, availability, scalability, and reliability of live services, with hands-on on-call/live-site experience, observability and monitoring, and unit/integration testing.
Won 10+ international hackathons (ETH India, ETH Bangkok) totalling $14,000+.
Spotmies
Remote
Decomposed a monolithic platform into 10+ microservices, improving availability and deployment reliability and enabling daily releases
Improved query performance by ~30% on high-traffic endpoints through PostgreSQL indexing and query analysis
Architected a fault-tolerant, event-driven Kafka pipeline sustaining 5K+ req/s with zero message loss, ensuring scalability under load
Cut mean time to detect from ~30 min to under 5 min via telemetry (structured logs, metrics, distributed tracing) and monitoring dashboards; mitigated live-site issues on a rotational on-call basis
Reduced redundant DB calls by 40% and p99 latency by ~60% with a Redis LRU cache; enforced per-tenant quotas via a token-bucket rate limiter
Raised test coverage to 85%+ by adding unit and integration tests with automated quality gates in CI/CD (Azure DevOps), cutting release turnaround to under 10 min
Built a RAG-powered knowledge assistant on Azure OpenAI, reducing support triage effort by ~35%
Reactive Network
Remote
Achieved sub-100ms p99 latency at 500 req/s on a secure finance API by optimizing SQL query execution and indexing, cutting DB processing time by 20%
Delivered 5 client MVPs end-to-end — architecture, testing, documentation, production deployment — with 100% on-time delivery in Agile sprints
Improved service health and scalability by identifying reliability risks and shipping architectural improvements that reduced operational overhead
B.Tech.
Java, Concurrency, Event Sourcing
Sustained sub-microsecond matching at 1M+ orders/sec in Java 21 via a single-threaded, lock-free, zero-allocation hot path (cache locality, branch prediction, fixed-point arithmetic)
Cut price-level access from O(logn) to O(1) and eliminated GC pauses using custom data structures (dense arrays, bitmaps, object pooling), verified via fuzz testing
Guaranteed RPO=0 crash recovery and exactly-once state reconstruction with write-ahead journaling and deterministic replay — core storage-engine and durability techniques used in databases
Java, Redis, React, MongoDB
Served 10K+ requests/day on a multi-tenant online service with role-based access control, tiered plans, and a token-bucket rate limiter
Improved response latency and reliability under load with an in-memory LRU cache for frequent RPC responses