Senior Software Engineer & Architect
Pushan Alagiya
Architecting scalable backend infrastructure, resilient distributed systems, and intelligent Agentic AI applications. Focused on sub-50ms latencies, high throughput, and robust production engineering.
Engineering Vision & Focus
With over 6 years of hands-on software engineering and architecture experience, I focus on designing distributed systems that sustain massive workloads while maintaining sub-50ms latencies and high reliability.
My technical writing and research cover Agentic AI orchestration (LangGraph, multi-agent systems, RAG with pgvector), asynchronous messaging pipelines with Apache Kafka, transactional outbox patterns, PostgreSQL internal mechanics, and high-performance full-stack web applications with React 19 and Next.js.
Academic Background & Honors
2024 - Present Helsinki, Finland
M.S. in Computer Science
University of Helsinki. Research and advanced coursework in Machine Learning, Distributed Algorithms, Large-Scale Data Systems, and Artificial Intelligence.
2020 - 2024 College Topper
B.E. in Information Technology
Government Engineering College, Bhavnagar (GTU). Graduated as College Topper with a CGPA of 9.45, demonstrating academic excellence across algorithms, distributed computing, database engineering, and networks.
Core Technical Disciplines
Agentic AI & LLMs
Multi-Agent Systems, LangGraph, LangChain, RAG Pipelines, Vector Search (pgvector, Pinecone), Tool Calling, Prompt & Context Engineering.
Backend & Distributed Systems
Node.js, NestJS, Express, Go, Microservices, Apache Kafka, BullMQ, Redis, PostgreSQL, MVCC, gRPC, WebSockets.
Full-Stack & Cloud
TypeScript, Next.js, React 19, Server Components, Tailwind CSS, Docker, Kubernetes, AWS, Serverless, Cloud-Native Architecture.
Featured Work & Projects
Distributed Collaborative Canvas
Low-latency real-time collaborative whiteboard utilizing WebSockets, CRDTs, and spatial R-tree rendering for 10,000+ interactive elements.
NVIDIA NIM Microservice Telemetry Monitor
GPU microservice telemetry collector monitoring token generation latency, GPU utilization, and throughput across model deployments.