Backend | Agentic AI | Distributed Systems
I build scalable backend systems and autonomous AI agents for high-traffic production environments.
From foundational IT education to production-grade backend systems
My journey began with a Diploma in IT, leading to a BE in IT where I developed a strong foundation in computer science. After securing a good rank in GATE (CS), I am currently pursuing postgraduate studies at IIT Bombay, focusing on deepening my understanding of distributed systems while continuing to build practical applications.
Focused on Node.js and Python for building scalable backend architectures. Experience ranges from real-time inventory systems to logistics optimization platforms and emergency response analytics.
I enjoy building systems that are robust and reliable. I strive to design architectures that can handle real-world challenges, such as unexpected traffic or system failures. I am always learning more about caching, async processing, database optimization and ways to improve API performance.
I approach problems by trying to understand the entire lifecycle of an application, from the initial API design down to deployment and monitoring, always looking for ways to improve overall observability.
I am currently exploring the frontier of Agentic AI, building autonomous agents and multi-agent systems that can reason, plan and execute complex workflows. I focus on making these systems reliable, observable and production-ready.
Production-ready skills organized by domain, not proficiency meters
Practical solutions to real-world engineering challenges
Jan'26 - Current
Monitoring critical events like wildfires, floods, etc. requires processing dynamic environmental data, predicting risks and making instant safety decisions.
Engineered a cohesive pipeline of 5 specialized agents (Data, Detection, Prediction, Decision, Alert) coordinated through a Flask backend. Built a real-time dashboard for interactive pipeline execution and alert management.
Created a functional multi-agent system capable of analyzing weather data, detecting anomalies and automatically triaging alerts.
May'25 - Jun'25
Mobile retail businesses faced critical inventory visibility and profit-tracking issues, leading to stock discrepancies and revenue loss.
Designed comprehensive backend workflows for sales, returns, supplier analytics and real-time stock valuation. Implemented Redis-based async processing to offload non-blocking operations and improve responsiveness.
Improved backend responsiveness through Redis-based async processing. Enabled real-time visibility into inventory and profit metrics for mobile retail operations.
Oct'25 - Dec'25
Manual truck allocation led to coordination overhead, delayed dispatch and inefficient route planning.
Built an Uber-style web application automating booking, fare estimation and end-to-end vehicle tracking with simulated GPS updates. Implemented traffic-aware routing and dynamic driver assignment using FastAPI-based REST APIs.
Reduced manual coordination overhead by automating booking workflows. Enabled faster dispatch and accurate fare estimation with interactive maps.
Oct'25 - Dec'25
Lack of visibility in ambulance operations resulted in delayed responses and inefficient route planning during emergencies.
Designed a GIS-based routing system with automated form tracking for trips. Implemented a robust 3-tier routing fallback system (ORS → OSRM → Haversine) to guarantee route calculation.
Delivered a reliable emergency routing platform. Improved operational visibility with animated map routing and data-driven response analytics.
Nov'25
High spectral redundancy in multispectral satellite imagery led to inefficient storage and processing.
Implemented Principal Component Analysis (PCA) mathematically from scratch. Built a Python GUI to visualize the principal components and reconstruct top-k images.
Successfully reduced spectral dimensionality while preserving information. Validated mathematical concepts by building the algorithm without high-level ML libraries.
Nov'25
Need for automated land cover classification from satellite imagery for environmental monitoring of the Powai Region.
Developed an unsupervised K-Means classification pipeline from scratch to identify Water, Vegetation and Others from Landsat images (2005-2025).
Successfully classified historical terrain data, enabling automated tracking of land/water changes over a 20-year span.
Technical deep-dives on Node.js, microservices and production systems
The ultimate, no-nonsense playbook for designing, building, and deploying scalable, production-ready microservices using Node.js.
A practical guide to building resilient microservices in Node.js by implementing advanced failure handling patterns, circuit breakers, and automatic retries.
Learn how to significantly improve your Node.js backend maintainability by adopting a structured, modular architecture pattern over the monolithic index.js.
Master the art of building scalable asynchronous messaging flows between microservices using event-driven architecture.
Unlock the full potential of Node.js by offloading heavy CPU-bound tasks to worker pools and distributed queues using BullMQ.
Protect your production backend systems from abuse, DDoS attacks, and traffic spikes seamlessly with advanced rate limiting and throttling strategies.
Production work, academic excellence and technical leadership
All India Rank 1814
Gujarat Technological University
Hacktober Hacks Hackathon & Hackout'21
L.J. Polytechnic
The Bharat Scouts & Guides
IIT Bombay
Placement Office, IIT Bombay
Self-Employed
Rapidops Inc.
HackVGEC'23 & TechXIT'22
IIT Bombay
Vishwakarma Government Engineering College
L.J. Polytechnic