Learning Roadmap#
Choose your learning path based on your goals. Each path builds on the previous one.
Quick Start#
Start here regardless of your goal:
| Step | Tutorial | What You'll Learn |
|---|---|---|
| 1 | Installation | Install Jac and plugins |
| 2 | Hello World | Basic syntax, functions |
Then build your first real app:
| Step | Tutorial | What You'll Learn |
|---|---|---|
| 3 | Part 1: Todo App | Nodes, def:pub, reactive frontend |
| 4 | Part 2: Add AI | by llm(), enum, obj, sem |
| 5 | Part 3: Walkers, Auth & Structure | Walkers, authentication, multi-file |
This 3-part tutorial covers graph storage, AI integration, walkers, and full-stack patterns in one cohesive project.
First App vs Deep-Dive Paths
The Build Your First App tutorial covers graph storage, AI, walkers, and full-stack patterns in one project. The paths below go deeper into each area individually. You don't need both -- choose based on whether you prefer learning by building or by topic.
After the first app tutorial, choose a deeper path:
Path A: Core Language (1 hour)#
Goal: Master Jac syntax and Object-Spatial Programming
| Step | Tutorial | Time | What You'll Learn |
|---|---|---|---|
| 5 | Jac Basics | 20 min | Variables, functions, objects |
| 6 | Object-Spatial Programming | 30 min | Nodes, edges, walkers in depth |
| 7 | Testing | 10 min | Write and run tests |
You can now: Write Jac programs, build graph-based applications, test your code.
Path B: AI Integration (1 hour)#
Goal: Build AI-powered applications with LLMs
Prerequisites: Quick Start + Path A recommended
| Step | Tutorial | Time | What You'll Learn |
|---|---|---|---|
| 5 | byLLM Quickstart | 15 min | First AI function |
| 6 | Structured Outputs | 20 min | Type-safe LLM responses |
| 7 | Agentic AI & Tools | 25 min | Tool calling, ReAct pattern |
You can now: Create AI-powered functions, build agentic systems, integrate LLMs.
Path C: Full-Stack Apps (2 hours)#
Goal: Build complete web applications with frontend and backend
Prerequisites: Quick Start + Path A recommended
| Step | Tutorial | Time | What You'll Learn |
|---|---|---|---|
| 5 | Project Setup | 15 min | jac-client, project structure |
| 6 | React-Style Components | 20 min | JSX in Jac, props |
| 7 | State & Effects | 20 min | Reactive state, hooks |
| 8 | Backend Integration | 25 min | Walker APIs, useWalker |
| 9 | Authentication | 20 min | Login, signup, protected routes |
| 10 | Routing | 15 min | Client-side navigation |
| 11 | Build a Todo App | 25 min | Complete example with AI |
You can now: Build full-stack web apps, handle authentication, manage state.
Path D: Production Deployment (30 minutes)#
Goal: Deploy Jac applications to production
Prerequisites: Quick Start + at least one other path
| Step | Tutorial | Time | What You'll Learn |
|---|---|---|---|
| 5 | Local API Server | 15 min | jac start, API endpoints |
| 6 | Deploy to Kubernetes | 15 min | jac-scale, K8s deployment |
You can now: Run production servers, scale with Kubernetes, deploy to cloud.
Complete Path#
For comprehensive understanding, follow all paths in order:
Quick Start + Build Your First App
↓
Core Language (deep dive)
↓
AI Integration (deep dive)
↓
Full-Stack Apps (deep dive)
↓
Production
Learning by Example#
After completing your chosen path, explore real-world examples:
| Example | Description | Skills Used |
|---|---|---|
| LittleX | Twitter clone | Full-stack, graph modeling |
| EmailBuddy | AI email assistant | AI integration, agentic patterns |
| RAG Chatbot | Document Q&A | AI, vector search |
| RPG Generator | Game level generator | Structured outputs, AI |
Quick Reference#
Need to look something up? Use the reference docs:
| Topic | Reference |
|---|---|
| Graph operations | Graph Operations |
| Walker responses | Walker Responses |
| Full language | Language Reference |
| CLI commands | CLI Reference |
| Configuration | Configuration |
| byLLM | byLLM Reference |
| jac-client | jac-client Reference |
| jac-scale | jac-scale Reference |
Stuck?#
- Troubleshooting - Common issues and solutions
- Discord - Ask the community
- JacGPT - AI-powered help