UpperCurve logoUPPERCURVE

6 Weeks Live Cohort · Limited Seats

Agentic AI
Launchpad

Go beyond just using AI, Learn to build AI Agents that actually Work.

6

Weeks Live Hands-On Program

10+

AI Agent Building

Certification Course

1:1

Mentorship Guidance

Enrolling now: Limited SeatsNext cohort

Go from Prompting to Deploying.

In six weeks, you’ll build practical AI Agents from scratch, work on real projects & develop the confidence to deploy what you create.

  • 10+ Hands-On Agent Projects.
  • Live Weekly Sessions with 1:1 Doubt Solving.
  • No Coding Experience Needed to Start.
CLAIM YOUR SEAT

Not ready? Try a Free Live Workshop

Every agent tool that matters. Covered.

Claude logoClaudeOpenAI logoOpenAIn8n logon8nMake.com logoMake.comCrewAI logoCrewAILangGraph logoLangGraphMCP logoMCPElevenLabs logoElevenLabsLovable logoLovablePython logoPythonSupabase logoSupabaseCursor logoCursorClaude Code logoClaude CodeHugging Face logoHugging Face

Everyone is talking to AI. Almost no one is Building with AI.

Learn to Build, Deploy and Ship production-ready AI agents in 6-Weeks.

The curriculum · 6 weeks · 6 modules

The deepest agent syllabus you can take live.

Every module ends with something running in the real world. Scroll through the stack — this is what the six weeks actually look like.

Module 01 · Week 1

INTRO TO AGENTIC AI & AI for work

CClaudeOOpenAI
What AI agents arePrompting & context engineeringLLM APIsAutomating PPTs, Docs & Excel
  1. 01.1How agents differ from chatbots: goals, tools, memory and loops
  2. 01.2Context engineering — structuring instructions that survive real inputs
  3. 01.3Calling LLM APIs directly, and choosing between models
  4. 01.4Automating the documents you already produce every week
Module 02 · Week 2

AI AGENTS FOR automation

nn8nMMake.comLLovableEElevenLabs
n8n workflowsMake.comVoice agentsVibe codingImage & video agents
  1. 02.1Designing trigger → reason → act workflows in n8n and Make.com
  2. 02.2A voice agent that answers, qualifies and books — end to end
  3. 02.3Vibe-coding a working product with Lovable
  4. 02.4Image and video generation pipelines you can hand to a team
Module 03 · Week 3

RAG & AGENTS from scratch

LLangGraphCCrewAIMMCPPPython
RAG & vector DBsChat with your dataTool callingMCP serversLLM evaluation
  1. 03.1Embeddings, chunking and retrieval that actually finds the answer
  2. 03.2Tool calling and agent loops without a framework, then with one
  3. 03.3Building and connecting your own MCP server
  4. 03.4Evaluating agents so you know when they are good enough to ship
Module 04 · Week 4

MASTERING THE Claude ecosystem

CClaudeAArtifactsCCowork
ArtifactsProjects & skillsAI memoryConnectors & MCPClaude Cowork
  1. 04.1Artifacts as shippable mini-apps, not just previews
  2. 04.2Projects, skills and memory for assistants that know your work
  3. 04.3Connectors and MCP: wiring Claude into calendars, files and apps
  4. 04.4Delegating real multi-step work with Claude Cowork
Module 05 · Week 5

CLAUDE CODE, DESIGN & multi-agent systems

CClaude CodeCCodexSSupabaseCCursor
Claude Code · CLI & webCodexAI designSupabase full-stackMulti-agent orchestration
  1. 05.1Shipping a full-stack app with Claude Code and Supabase
  2. 05.2Design-to-code workflows that keep a product looking intentional
  3. 05.3Orchestrating specialist agents that hand work to each other
  4. 05.4Deploying to a real domain, with auth and a database
Module 06 · Week 6

CAPSTONE: launch your agent

11:1 mentorshipDDemo Day
Your idea1:1 reviewsDemo DayDeployed & public
  1. 06.1Scope an agent around a problem you or your team really have
  2. 06.2Two 1:1 reviews to pressure-test the design and the build
  3. 06.3Present it live on Demo Day to mentors and the cohort
  4. 06.4Leave with it deployed, public and in your portfolio

What you will build

Six weeks. Fifteen things that work.

Not exercises. Not notebooks. Agents that keep running after the session ends.

Build 01

A website, live by dinner.

Your idea on a real domain in one sitting — designed, coded and deployed with Claude Code. Frontend, database and auth, done.

Claude CodeSupabase

Build 02

A research team that never sleeps.

A researcher, an analyst and a writer: three agents that search, verify and hand you a finished brief while you are in a meeting.

CrewAILangGraph

Build 03

A voice agent that picks up.

It answers calls, qualifies the lead and books the slot on your calendar — then logs the conversation where your team can see it.

ElevenLabsn8n

Build 04

Chat with your own data.

A retrieval agent over your documents that cites its sources, so the answers are ones you can actually forward.

RAGVector DB

Build 05 · Capstone

Your personal chief of staff.

Claude connected to your calendar, files and apps through MCP. One assistant that plans your week and preps your meetings — then Demo Day.

MCPClaudeEverything above

Drag to explore builds

How the cohort runs

Built live, reviewed 1:1, shipped in public.

Taught by practitioners who ship agents at work — not presenters. Every week has the same rhythm, so building becomes a habit rather than a heroic weekend.

  1. 01Every week

    Live build sessions

    Two evenings a week, building alongside the mentor — not watching slides.

  2. 02Every week

    Build labs & office hours

    Unblock your project in small groups before it turns into a lost weekend.

  3. 03Weeks 5–6

    1:1 capstone reviews

    Your design and your build, pressure-tested one-on-one before it goes live.

  4. 04Week 6

    Demo Day

    Present a deployed agent to mentors and the cohort. It goes straight into your portfolio.

Where this takes you

These roles barely existed two years ago. They are hiring now.

Finish the cohort with a portfolio of working agents and you are a credible candidate for the fastest-growing roles in tech.

Hiring signal · live

  1. 01

    Forward deployed engineer

    Customer-facing

    Embed with customers and build agent solutions on-site. One of the fastest-growing titles in AI.

  2. 02

    AI product manager

    Product

    Ship AI products with the judgement of someone who has built agents, not just read about them.

  3. 03

    Agent engineer

    Engineering

    Design, build and maintain production agent systems — the core engineering role of this decade.

  4. 04

    AI automation consultant

    Independent

    Help teams replace repetitive work with automations they can trust and maintain themselves.

Proof you can show

Get recognised for your agentic expertise.

Finish the cohort and earn a certificate backed by real shipped work, not watch-time. Put it on LinkedIn next to the projects it certifies.

  • Awarded on completion — tied to your Demo Day project
  • A unique ID anyone can verify
  • Backed by a portfolio of fifteen projects employers can open

UpperCurve · Credential

UPPERCURVE

Certificate in Agentic AI Engineering

This certifies that

Your Name Here

has completed the six-week Agentic AI Launchpad, shipped fifteen agentic projects and presented a production agent at a live Demo Day.

Programme Director

UpperCurve

Verified

Unique credential ID

Your seat

An investment in who you become next.

This isn't a course purchase. It's a decision about which side of the AI divide you stand on six weeks from now.

  • 6 weeks of live sessions with recordings of every one
  • 15 guided projects including your own capstone agent
  • Ready-made agent templates for Claude Code and Codex, yours to reuse at work
  • 1:1 capstone mentorship and a live Demo Day
  • Private builder community with weekly office hours
  • Certificate in Agentic AI Engineering shareable on LinkedIn
Next cohort

One seat · everything included

Limited seats.

Fees and batch dates are shared when you apply — so 1:1 mentorship stays 1:1.

6 weeks of live training15 hands-on projects1:1 capstone mentorshipCertificate for LinkedIn
Apply for the cohort

Still deciding? Attend a free live workshop first.

Asked before enrolling

The questions everyone actually asks.

Who is this cohort designed for?

Professionals, founders, product people and developers who want to build with AI rather than just use it. You do not need to be an engineer — the first two modules are no-code, and code is introduced gradually.

Do I need coding experience?

No. You start with no-code automation tools and move to code with AI coding assistants doing the heavy lifting. If you already code, the later modules go deep enough to stay interesting.

How much time does it take each week?

Plan for six to eight hours: the live sessions, plus time on that week's build. Build labs and office hours are there so you are never stuck for long.

What will I have at the end?

Fifteen working projects, a deployed capstone agent you presented at Demo Day, and a certificate backed by that portfolio.

What if I miss a live session?

Every session is recorded and shared the same week. Office hours are the place to catch up on anything you missed.

Is there a certificate?

Yes — the Certificate in Agentic AI Engineering, awarded on completion of the capstone. Every certificate carries an ID anyone can verify.

How do I get in touch with questions?

Join the UpperCurve community and ask there, or come to one of our free live events first to see how we teach.

Next cohort · Seats limited

Stop prompting.Start shipping.

The next cohort starts soon. Seats are capped so every capstone gets a real 1:1 review.

Join the next cohort