I'm a third-year Integrated M.Tech student in Information Technology at ABV-IIITM Gwalior, focused on backend engineering, systems programming, distributed systems, and AI infrastructure.

I enjoy working close to the system boundary, from low-latency C++ and concurrency to cloud-native infrastructure and scalable AI systems.

Techneow
January 2026 – July 2026

• Developed and enhanced backend services during a 6-month internship, focusing on performance, reliability, and maintainable software design.
• Optimized backend operations by identifying inefficient implementations and improving system-level performance and execution efficiency.
• Engineered reliable backend components by investigating technical issues, strengthening failure handling, and improving application stability.

CircuitEvolve
August 2026 – September 2026

• Engineered and debugged a Cadence Virtuoso-based GUI workflow, integrating simulation, schematic generation, and automated validation across 900+ tests.
• Automated end-to-end circuit simulation workflows, reducing repetitive manual setup and validation across multiple Cadence execution paths.
• Improved GUI reliability and regression coverage across 4+ major workflow components, resolving waveform/PSF extraction, environment, and process-launch failures.

Contributor to CloudNativePG, a Kubernetes operator for PostgreSQL.

• Extended CloudNativePG's Kubernetes API to support dnsPolicy and dnsConfig, propagating native PodDNSConfig across 2 workload paths: PostgreSQL instance Pods and operator-managed Jobs, with CRD, API, deepcopy, and unit-test updates.
• Fixed PodMonitor lifecycle reconciliation to correctly clean up Pooler-owned monitoring resources when monitoring is disabled, while preserving safeguards for manually managed resources and adding regression coverage.
• Resolved upstream integration conflicts across generated APIs, CRDs, Pod/Job specifications, and reconciliation logic while maintaining passing CI and regression tests.

Real-time industrial safety and emergency response platform combining deterministic risk detection with AI-assisted decision support.

Tech: FastAPI • WebSockets • Next.js • RAG • Neo4j • Multi-Agent AI

• Engineered a real-time safety platform across 20+ plant areas using FastAPI, WebSockets, Next.js, and deterministic compound-risk detection.
• Built a 5-agent AI emergency assistant integrating RAG, Neo4j, intent routing, and predictive analysis for grounded operator decisions.
• Achieved 363 req/s, 8.9 ms P95 latency, 0% errors, with 7.9 ms P95 telemetry-to-alert propagation under a 20-VU workload.

Ultra-low-latency limit order book and matching engine built in C++20.

Tech: C++20 • Multithreading • Lock-Free Concurrency • Memory Management

• Achieved ~730 ns average matching latency while stress-testing the matching engine with 1M+ synthetic orders.
• Built a lock-free SPSC queue achieving 7.7M+ orders/sec throughput, eliminating lock contention from the critical data path.
• Reduced runtime heap-allocation jitter by 99.9% using a custom slab allocator combined with CPU core thread affinity.

Real-time collaborative coding and communication platform supporting synchronized editing and video communication.

Tech: MERN • Socket.IO • WebRTC • JWT

• Supported 50+ concurrent user sessions using WebSocket-based real-time state synchronization.
• Reduced video streaming bandwidth by 80% using WebRTC peer-to-peer communication.
• Achieved sub-100 ms synchronization latency for collaborative sessions.

In Progress

Distributed infrastructure for scalable LLM inference across multiple worker nodes.

Tech: Python • FastAPI • Redis • PyTorch • Hugging Face • Docker

• Designing an asynchronous inference pipeline using FastAPI, Redis, and distributed workers to decouple request ingestion from model execution.
• Exploring tensor parallelism, continuous batching, and KV-cache optimization to improve GPU utilization and inference throughput.
• Building the architecture around measurable latency, throughput, memory-efficiency, and horizontal scalability benchmarks.

• Google Big Code 2026: Ranked Top 1500 globally.
• Competitive Programming: Solved 400+ DSA problems across LeetCode, Codeforces, and CodeChef.
• JEE Mains: Scored 99th percentile among 1.4M+ candidates.
• Smart India Hackathon: Led a 5-member team to build a full-stack agricultural e-commerce marketplace.