← Back to CABColumbus App BuildersColumbus App BuildersAcademy
Apply

Agentic Systems & Application Engineering

Design. Orchestrate. Build. Deploy.

Learn to design, build, evaluate, and deploy intelligent software systems that use AI agents, tools, data, automation, durable state, and modern application architecture. Prework, 8 core weeks and a Week 9 of capstone presentations; about 6–10 hours per week.

Apply or join the waitlist

Syllabus

  1. Prework: Building Blocks

    You will learn

    • how modern software applications work
    • frontend vs backend
    • APIs
    • databases
    • authentication
    • JSON
    • basic JavaScript/TypeScript concepts
    • basic Python concepts

  2. Week 1: From AI Model to AI Application

    You will learn

    • LLM vs application
    • generative AI vs agentic AI
    • model limitations
    • context
    • structured outputs
    • API-driven AI
    • system instructions
    • tool calling

    You will demonstrate

    • Core Principle — AI HANDLES JUDGMENT.
    • Lab — Build a basic AI-enabled application that:
    • Mission — Identify which parts of a business workflow should and should not use AI.

  3. Week 2: Building Powerful Applications

    You will learn

    • product architecture
    • user roles
    • permissions
    • databases
    • Supabase
    • authentication
    • server-side functions
    • APIs

    You will demonstrate

    • Lab — Build a small full-stack application with:

  4. Week 3: Agents, Tools & Reasoning Loops

    You will learn

    • what an agent actually is
    • goal
    • state
    • tools
    • observation
    • action
    • iteration
    • termination

    You will demonstrate

    • Lab — Build an agent using several tools.

  5. Week 4: Knowledge, RAG & Context Engineering

    You will learn

    • embeddings
    • vector databases
    • chunking
    • metadata
    • retrieval
    • hybrid search
    • semantic retrieval
    • permissions

    You will demonstrate

    • Lab — Build a domain-specific knowledge agent.

  6. Week 5: Agent Skills & Multi-Agent Orchestration

    You will learn

    • Skill metadata
    • discovery
    • activation
    • scope
    • inputs
    • outputs
    • resources
    • procedures

    You will demonstrate

    • Lab — Convert one overloaded agent into:

  7. Week 6: Plan → Execute → Evaluate & Parallelism

    You will learn

    • independent critics
    • quality gates
    • reflection
    • bounded correction loops
    • pass/fail criteria
    • model-as-judge
    • deterministic validation
    • human escalation

    You will demonstrate

    • Lab — Build a generator + critic system.

  8. Week 7: MCP, Interoperability & Durable State

    You will learn

    • tools
    • resources
    • prompts
    • connectors
    • agent-to-agent communication

    You will demonstrate

    • Lab — Connect an agent to external capabilities through a protocol-compatible architecture.

  9. Week 8: Evals, Security & Production Deployment

    You will learn

    • traces
    • run IDs
    • logs
    • tool-call history
    • latency
    • tokens
    • model cost
    • failures

    You will demonstrate

    • Lab — Students create an evaluation suite for their agent.

  10. Week 9: Capstone Presentations

    You will learn

    • construction operations
    • creator command center
    • customer intake
    • field dispatch
    • research system
    • sales qualification
    • education platform
    • agentic project management

CAB Academy is a professional training program of Columbus App Builders. It is not accredited, does not grant degrees or academic credit, and is not endorsed by any technology company or university.