Agentic AI
Architect

From prompts to autonomous systems with MCP & Spec-Driven Development. Two days. One live system: the tools an AI uses, then the app that displays what it wrote.

AICTE Indovation Centre, Jaipur — Technovate Hub  ·  GITS, Udaipur  ·  June 18–19, 2026

The one idea

Building with AI

Using a chatbot as a smarter autocomplete. Prompt, copy, paste, hope. You consume tools someone else built.

Building AI systems

Designing the tools an agent calls, the harness it runs in, and the spec that governs the code. You build the machine — and ship it live.

The two-day arc

Day 1 · The MechanismBuild the tools an AI usesAgent anatomy → MCP architecture → build your own MCP server (StackLog) → wire it into a multi-tool agent pipeline, then break it on purpose and harden it. 11 modules. Day 2 · The MethodSpec & build the appSDLC→AI-DLC → why specs beat prompts → the SDD lifecycle with BMAD V6 → spec, decompose, generate, and review the web app that displays the agent's data. 8 modules.
Audience
Final-year eng students, developers, founders
Assumed
JavaScript / TypeScript + Git. No prior AI-agent experience.
Capstone
StackLog — one prompt, two days, one running system
Environment
Google Antigravity (IDE + agent)
Start Day 1 →Resources