name : Shubh Gupta
alias : titanShubh
college : IET Lucknow, B.Tech CSE (2023–2027)
focus : ["GenAI Engineering", "Multi-Agent Orchestration", "SDE", "Competitive Programming"]
building : Multi-agent AI systems, RAG pipelines, LLM-powered apps
ask_me_about: ["LangChain", "LangGraph", "RAG", "Vector DBs", "DSA", "System Design"]
reach_me_at : [LinkedIn, LeetCode, Codeforces, CodeChef, AtCoder]
fun_fact : "I treat every CP problem like a production bug 🐛"
status : "Actively seeking GenAI & SDE internships/roles 🚀"🤖 AI / GenAI & Agentic Systems
🗄️ Vector DBs & Storage
⚙️ Languages & SDE
🗃️ Databases & Caching
🛠️ Tools, Data & DevOps
| Platform | Profile | Rating & Level |
|---|---|---|
| titan_2 | 🔵 Expert (1794) | |
| guptashubh6386 | 🟡 Guardian (2245) | |
| shubh_titan | 🔶 5-Star | |
| unknown_AC | 🔵 2 Kyu Blue (1617) |
Solved 2500+ algorithmic problems across platforms
- 🏆 Global Rank 38 in CodeChef Starters 169 (Rated, Division 2)
- 🏆 Global Rank 199 in Codeforces Round 1029 (Div. 3)
- 🏆 Global Rank 358 in LeetCode Weekly Contest 422 (26,000+ participants)
- 🏆 Global Rank 2881 in Meta Hacker Cup 2024 (20,000+ participants globally)
Architected a stateful multi-agent system utilizing LangGraph and LangChain.
- Deployed a central Supervisor Agent that dynamically routes queries across seven specialized nodes (Problem Analyzer, Complexity Analyzer, Learning Memory, etc.).
- Engineered an asynchronous FastAPI backend using SQLModel and Server-Sent Events (SSE) to stream structured reasoning logs with sub-100ms latency.
- Configured Qdrant Cloud Vector Database for semantic problem retrieval integrated with PostgreSQL for persistent tracking.
A high-performance stateful multi-agent system designed for automated query routing.
- Utilized LangGraph to analyze query intent and orchestrate execution across specialized SQL, RAG, and Python Pandas nodes.
- Built a secure, self-healing SQL executor using SQLAlchemy 2.0 with an AST-based validator to prevent injections.
- Built a spreadsheet analytics sandbox utilizing Pandas and Plotly to render statistical charts dynamically.
A multi-threaded port scanner and real-time vulnerability analysis engine.
- Engineered a socket-level host discovery engine using ThreadPoolExecutor in Python.
- Implemented bidirectional WebSockets communication to stream live scan telemetry and latencies to a React dashboard.
- Containerized the application stack (FastAPI, Nginx, SQLite, React) using Docker Compose.
⚡ Auto-updates daily via GitHub Actions