M.Sc. Student in Artificial Intelligence at the University of Jyväskylä.

I bridge the gap between AI Research and Production Engineering. With a background in high-performance computing and simulation (8+ years), I now architect secure, scalable AI Infrastructure. I specialize in building Agentic Workflows and Enterprise RAG Systems that prioritize governance, safety, and reliability over simple "demos."

The Pitch: A production-grade Enterprise RAG Platform designed for SMEs and regulated industries.

• Core Tech: Python (FastAPI), LangGraph, Azure OpenAI, Docker, PostgreSQL (pgvector).
• Key Features: Implements Agentic Intent Routing to distinguish between chat and retrieval, significantly reducing token costs. Enforces Strict Citation Governance using Pydantic validation to eliminate hallucinations.
• Architecture: Fully containerized Microservices architecture ready for Azure deployment.

The Pitch: A next-generation AI-powered CLI Assistant that acts as a "Copilot for your Terminal."

• Core Tech: Go (Golang), Local LLM Inference (Llama 3), Custom JSON Planner.
• Key Features: Bridges natural language and system automation using a custom Safety Sandbox layer. Demonstrates deep systems programming and memory optimization in Go.

The Pitch: An Open-Web Retrieval Engine for AI pipelines (Open Source alternative to paid Search APIs).

• Core Tech: Go, SmartQuery Routing, Robots.txt Compliance Engine.
• Key Features: Delivers high-performance web scraping with metadata normalization and RSS/Atom parsing. Features a plugin system for modular data extraction.

The Pitch: A Full-Stack Academic Assistant enabling personalized learning paths.

• Core Tech: React (Vite), TypeScript, Go (Fiber), RAG.
• Key Features: Showcases end-to-end product engineering with a reactive Streaming Chat Interface and local inference orchestration.

• Frameworks: Python (Advanced), LangChain, LangGraph, Pydantic AI, LlamaIndex.
• LLM Ops: Azure OpenAI, Ollama (Local), Prompt Engineering, RAGAS Evaluation, CI/CD for AI.
• Agentic Patterns: Tool Calling (MCP), Multi-Agent Orchestration, Structured Output Validation.
• Vector Search: Hybrid Retrieval, Re-ranking (Cohere/Cross-Encoders), Qdrant, Milvus.

• Backend: FastAPI (Python), Fiber/Gin (Go), Microservices, REST & gRPC.
• Frontend: React, TypeScript, TailwindCSS, Vite.
• DevOps: Docker, Kubernetes, GitHub Actions, Azure Container Apps.
• Data: PostgreSQL (pgvector), Redis.

• Simulation: Real-time Systems, Digital Twins, C++ Optimization.
• Visualization: 3D Systems, Rendering Pipelines (Unreal Engine).

• LinkedIn: linkedin.com/in/nibir-1
• Email: [email protected] | [email protected]