JanSetu AI
AI-powered citizen welfare gateway enabling vernacular voice and text access to over 500+ Indian public welfare programs and legal provisions.
PORTFOLIO ARCHIVE
CSE STUDENT • DEVELOPER • AI BUILDER

Building scalable AI backend architectures, high-performance APIs, and intelligent agent systems with engineering discipline.
Full Stack Web & AI Systems
36-Hour Rapid Prototyping
Java, Next.js, Gemini, RAG
Algorithmic & Architecture Rigor
Open Source & Experimental Works
Exploring Frontier Technology
An editorial selection of full-stack platforms, multimodal AI systems, and low-latency tools engineered with an obsessive focus on performance and architectural clarity.
AI-powered citizen welfare gateway enabling vernacular voice and text access to over 500+ Indian public welfare programs and legal provisions.
Specialized regulatory intelligence assistant parsing tens of thousands of pages of Indian Bureau of Standards (BIS) industrial specifications.
Dense BGE embeddings combined with sparse BM25 term weighting, cross-encoder re-ranking, and clause-level citation linking.
Dynamic student digital twin engine architected by Rudra as CTO & Tech Lead at DTV, simulating campus placement trajectories and generating personalized roadmap graphs.
Deterministic gap analysis matching academic milestones, DSA problem vectors, and system architecture benchmarks.
ENGINEERING AT THE INTERSECTION OF RIGOR & CREATIVITY
HOVER OR SELECT A PHASE TO INSPECT:
Designing high-throughput asynchronous APIs, FastAPI services, and distributed cloud backends with rigorous latency guarantees.
Directed core technical team and web infrastructure at Digital Twin Verse, scaling immersive 3D digital twin systems.
Pioneering Kaggle AI agent workflows, Google Gemini reasoning loops, and multi-source RAG architectures.
First-principles academic and algorithmic foundations in Data Structures, Java, and System Architecture.
Synthesizing robust systems programming with editorial aesthetics, micro-interactions, and real-time experiences.
PROFESSIONAL SYNOPSIS
As a Computer Science Engineering student at UIT RGPV and former Chief Technology Officer at Digital Twin Verse (DTV), I don't just build interfaces—I architect robust backend foundations, optimize complex data structures, and harness modern Large Language Models to solve meaningful problems.
My engineering approach prioritizes speed, modularity, and intentional design over bloated frameworks. From civic platforms like JanSetu AI to high-concurrency FastAPI microservices and digital twin systems, every architecture is crafted to deliver palpable utility with editorial restraint.
I evaluate competency not through arbitrary percentage meters, but by systems shipped, algorithms implemented, and latency benchmarks achieved.
DSA, Multithreading, OOP, JVM Architecture, LeetCode 500+
Strict static types, Generics, Next.js App Router, SDK Dev
ESNext, Asynchronous Event Loops, WebGL & DOM APIs
FastAPI, LangChain, PyTorch embeddings, Data Wrangling
Tangible laurels from national hackathon stages, certified cloud architectures, and algorithmic problem-solving benchmarks.
Official Google Cloud skill badge demonstrating architectural competency across Google Cloud infrastructure, Vertex AI, foundation models, and prompt orchestration.
Engineered complex prompt pipelines and multi-agent reasoning loops under competitive sprint constraints in the Google for Developers Build with AI arena.
Honored for exceptional technical leadership as Chief Technology Officer, directing web infrastructure, scalable frontend architectures, and UI/UX design systems.
Rigorous immersion in autonomous AI agents, tool-calling loops, multi-agent collaboration, memory stores, and rapid generative prototyping.
Designed scalable civic technology prototypes targeting public governance, high-velocity community accessibility, and demographic personalization.
Shortlisted among top engineering teams at UIT RGPV to represent the institute for Smart India Hackathon problem statements solving civic bottlenecks.
Competed at national finals in building multi-agent coordination systems, autonomous tool invocation routines, and low-latency agent workflows.
Rigorous computational problem solving covering Dynamic Programming, Graph Theory, Segment Trees, Backtracking, and Asymptotic Optimization in Java.
Organized and coordinated institute-wide coding event promoting algorithmic problem solving, peer mentorship, and competitive programming culture.
From first-principles theoretical computation to national hackathon podiums, multimodal AI deployments, and creative web engineering.
Data Structures, Java Rigor, Algorithms & Systems Architecture
Established first-principles algorithmic foundations, solving complex computational problems with space-time optimality.
Autonomous Agent Orchestration, Civic Bottlenecks, Rapid Prototyping
Selected in UIT RGPV internal hackathon for SIH 2025 and awarded Google for Developers Agentic Premier League Finals achievement.
Algorithmic Problem Setting, Technical Volunteering, Peer Mentorship
Active member organizing technical operations, problem curation, and coding sprints for 150+ student participants.
Google Cloud Infrastructure, Vertex AI, Foundation Models, Prompt Engineering
Earned official Google skill accreditation (Badge ID: 22452714) demonstrating enterprise generative AI proficiency.
Next.js App Router, Responsive UI/UX, Real-Time Dashboards, API Integration
Engineered student digital twin simulation engines and responsive analytics interfaces for university recruitment.
Core Tech Leadership, Architecture Governance, UI/UX System Design
Directed engineering team and product infrastructure, honored with official Certificate of Recognition (DTV-CORE-2026-006).
Google Build with AI, Kaggle 5-Day Vibe Coding, Unstop Civic Innovation
Concurrently completed national competitions with verified credentials from Google for Developers, Kaggle, and Unstop.
FastAPI, Hybrid RAG, Autonomous Tool-Calling Agents, Low-Latency WebSockets
Architecting high-throughput asynchronous backend systems and multimodal AI workflows with rigorous engineering discipline.
Whether you are looking to collaborate on high-impact AI products, build cutting-edge web platforms, explore hackathon ventures, or discuss computer science architecture—my inbox is always open.
NETWORK DIRECTORY