[01]
AI Agent Systems
Orchestrating multi-agent systems that model software architecture and generate full-stack APIs.
FastAPI · MongoDB · LangGraph · Google ADK · React
Impact28_AGENT_WS
I build scalable backend systems, AI-powered products, and machine learning infrastructure that solve real-world problems. Focused on auditability and fault tolerance.
I am a Computer Science & Engineering student at Parul University (Aug 2023 – Apr 2027) specializing in high-throughput backend architecture, distributed systems, and Machine Learning Operations (MLOps).
My passion lies in bridging reliable, production-grade infrastructure with intelligent systems. I design architectures that remain resilient under real-world load, optimize queries, and leverage machine learning to automate complex tasks.
“Software engineering is about writing clean, auditable, maintainable, and self-documenting systems that remain reliable under real-world scale.”
B.Tech Computer Science & Engineering
August 2023 – April 2027
Backend, Spring Boot, event systems
Models, MLOps, deep learning
Academic profile, general systems
Orchestrating multi-agent systems that model software architecture and generate full-stack APIs.
FastAPI · MongoDB · LangGraph · Google ADK · React
High-throughput microservices using asynchronous queue ingestion and caching.
Java · Spring Boot · Kafka · Redis · Docker
Stateful WebSocket synchronizers supporting concurrent drawing, canvas versions, and socket nodes.
React · Node.js · Express · MongoDB · Redis · Socket.io
ACID-compliant transactional routers enforcing FEFO shelf-life models.
React · Node.js · Express · MongoDB
Two-Tower deep learning model pipelines for fast semantic retrieval.
Python · TensorFlow · FAISS · SentenceTransformers · FastAPI
Ensemble model evaluators integrating SHAP contributions for fully auditable valuation outputs.
Python · XGBoost · SHAP · FastAPI · Docker
Tokenizers and classifier pipelines analyzing sentiment structures and political article bias.
Python · FastAPI
Continuous integration loops checking for data drift, tracking metrics, and publishing models.
Python · MLflow · DagsHub · FastAPI · MongoDB · Docker
Enterprise-grade AI Co-Founder workspace that transforms product concepts into implementation-ready software systems using a cyclic multi-agent graph.
Natural language requirement ingestion engine.
Orchestration flow & transition diagrams.
Secure MongoDB schema & index mapping.
Full-stack FastAPI models and UI structures.
Validations, syntax repair, and path refactors.
Pipeline foundation
<10ms API overhead
Multi-threaded services
Event-driven sockets
Production core runtime
Reactive interfaces
28-agent graph
Self-healing compiler
R² = 0.847
Sub-10ms retrieval
Feature attribution
Two-Tower retriever
Sub-5ms cache
ACID schemas
Optimized queries
High-volume streams
Hermetic builds
Shell automation
Model registry
DVC data tracking
Demonstrated solid understanding of algorithms, data structures, complexity analysis, and problem-solving paradigms on LeetCode.
Competed against 700+ teams, designing and pitching a high-performance solution under intense time pressure.
Coordinated development sprints, architectural reviews, and deployment pipelines for collaborative team projects.
Designed and deployed Saarthi, integrating multiple specialized LLM agents with local filesystem tools and databases.
Writes in-depth technical case studies and architectural breakdowns to share distributed systems knowledge.
“Software engineering is about building systems people can trust.”
Auditability. Scalability. Fault Tolerance.