👋 Hi, I'm Harnoor Singh

AI Full Stack Engineer • Agentic Systems • RAG • Scalable Backend

I build production AI systems, intelligent agents, RAG applications, and high-volume backend platforms.

4+ years of software engineering experience across fintech, AI products, distributed systems, and full-stack development.

I'm an AI & Full Stack Engineer based in India 🇮🇳, focused on building intelligent products that work reliably in production.

At Twinline Business Solutions, I work across AI systems, fintech backend infrastructure, APIs, databases, and customer-facing applications.

My work sits at the intersection of:

AI Agents • RAG • Backend Engineering • Full Stack Development • Distributed Systems

I enjoy taking systems from architecture → implementation → evaluation → production.

At work, I design and build AI agents for collections field-force planning that generate validated:

• Monthly plans
• Weekly plans
• Daily visit schedules
• Optimized operational workflows

The system combines LLMs, planning logic, validation layers, backend services, and human workflows to replace previously manual planning processes.

AI-powered language test preparation platform serving 3,000+ students

🌐 aimseducation.org

Built an AI-driven platform for TCF Canada, IELTS, PTE, and CELPIP preparation with:

• AI speaking and writing evaluation
• Automated LLM feedback
• Speech-to-text transcription
• Timed practice and mock tests
• Student progress tracking
• Evaluation pipelines for language proficiency

Tech: Node.js Express PostgreSQL Supabase Groq ElevenLabs LLMs

Conversational RAG platform for querying documents and websites

🌐 mindbook-rho.vercel.app

Designed an end-to-end RAG system with:

• Document and website ingestion
• OCR processing
• Intelligent chunking
• Embedding generation
• Vector retrieval
• LangGraph agent orchestration
• Asynchronous background processing

Tech: Next.js TypeScript FastAPI Python LangChain LangGraph Celery Redis AWS

Multi-agent AI investment research and risk profiling platform

🌐 invisly.vercel.app

Built an LLM-powered investment analysis system with specialized agents for:

• 📰 Market news research
• 📊 Fundamental analysis
• 📉 Technical analysis
• 🧠 Risk profiling
• ✉️ Automated research reports

Tech: Next.js TypeScript LangChain Gemini PostgreSQL MongoDB Inngest

RAG-powered developer collaboration platform over GitHub repositories

🌐 gitlic.vercel.app

Built an end-to-end retrieval system allowing engineering teams to query:

• GitHub repositories
• Pull requests
• Commits
• Repository history

using natural language.

Tech: Next.js TypeScript tRPC Prisma PostgreSQL Gemini LangChain AssemblyAI

I actively contribute to production-grade AI and developer infrastructure projects.

Open-source AI agent

Contributed multiple merged pull requests across the desktop application and UI, including work around:

• Provider search and filtering
• Configuration Editor reliability
• Local inference model configuration
• Session deep-link handling

Contributed per-session LLM token budgeting, including hard limits and warning thresholds to prevent uncontrolled context spending.

Contributed a fix for Gemini thought-signature persistence across multi-step tool calls, improving correct multi-turn agent behavior.

RAG • AI Agents • Multi-Agent Systems • Vector Search • Embeddings • Reranking • Prompt Engineering • MCP • A2A • LangSmith • DeepEval

Python TypeScript JavaScript Java SQL

Node.js Express FastAPI Spring Boot Hibernate Celery REST APIs Microservices

PostgreSQL pgvector MongoDB Redis Oracle Prisma Supabase

AWS Docker CI/CD Vercel Cloudflare Git GitHub

AI Engineering
├── Agentic Systems
├── RAG & Retrieval
├── Multi-Agent Workflows
├── LLM Evaluation
├── Tool Calling
├── Memory & Context Management
└── AI Observability

Backend Engineering
├── High-throughput APIs
├── Distributed Systems
├── Microservices
├── Database Optimization
├── Caching
├── Async Processing
└── Production Reliability

Full Stack
├── Next.js
├── React
├── TypeScript
├── Node.js
├── FastAPI
└── PostgreSQL

I'm interested in building AI-native products, agentic systems, RAG platforms, developer tools, and scalable backend infrastructure.

If you're building something ambitious around AI or distributed systems, I'd love to connect.