🥇 1st Prize, Best Use of AI — TCS Tech Day 2026// 2026• ai
Bag Packer
NLP Catalogue Search & MCP Server Engine
Natural-language catalogue search and luggage packing system. Built a custom Model Context Protocol (MCP) server so Claude or any MCP-compatible app can directly query catalogue listings. Won 1st Prize in the Best Use of AI category at TCS Technology Day 2026.
OPINIONATED ARCHITECTURAL DECISION
Implemented Model Context Protocol (MCP) server endpoints natively so third-party AI agents (Claude Desktop, etc.) can directly invoke search tools.
WHAT I GOT WRONG & LESSONS LEARNED
First prototype passed entire catalog chunks into prompt context instead of structured tool filtering, exhausting token windows on large queries.
How I built this & what it delivers
Led the team building Bag Packer and won 1st Prize in the Best Use of AI category at TCS Technology Day 2026.
Implemented natural-language search over travel and packing catalogues.
Built a custom Model Context Protocol (MCP) server enabling Claude and external agents to query listings programmatically.
Open-source repository available at github.com/ishu17077/bag_packer.
// TECH ARSENAL
PythonFastAPINLPModel Context Protocol (MCP)LLM Tool CallingREST API
// PERFORMANCE & IMPACT
🥇 1st Prize, Best Use of AI @ TCS Tech Day 2026 • Custom MCP Server
Audited for production reliability, memory efficiency, and execution safety.