Ai 101 - What Is AI - Part 2
Building a “Jarvis”-Style AI System
RAG · Voice AI · Automation (In Person, SF)
In this in-person workshop at AI Start Academy SF, we go beyond prompts and demos and break down how real AI assistants are actually built.
You’ll see — step by step — how to design a “Jarvis”-style AI system that can:
Understand requests
Retrieve knowledge reliably
Speak back with a natural voice
Execute logic and actions
This is not a lecture.
It’s a system-level walkthrough with a live build.
What we’ll build (live)
A practical AI assistant using:
n8n for orchestration and logic
RAG (Retrieval-Augmented Generation) for grounded responses
ElevenLabs for high-quality voice output
Architecture overview:
Voice → LLM → RAG → Logic → Voice Response
We’ll focus on why the system is designed this way — not just how to connect tools.
What you’ll learn
How RAG makes AI assistants reliable and factual
The difference between AI agents and simple workflows
How n8n fits into modern AI automation stacks
Where voice AI belongs in real products
Key trade-offs: latency, cost, reliability, and scalability
How to reason about AI architecture before you build
Who this event is for
Founders and operators building with AI
Engineers and technical builders
Product, design, and ops leaders
Anyone who understands LLM basics (or attended Part 1)
Not beginner-level prompting.
This is for people ready to build systems.
Format
In-person workshop
Live system walkthrough
Q&A and open discussion
Small group, high signal
Location
AI Start Academy SF
Mission District, San Francisco