EkenLab
Open app

AI COURSE ADVISOR

What would you like to learn?

Chat with our AI advisor to find your first practical AI course.

Free AI chat · No sign-up needed

About this AI chat & your data

Up to 10 guest messages. OpenAI processes your messages. EkenLab keeps this conversation, your stated preferences and basic device details (browser, device type, country) in its analytics archive to improve courses; precise IP, city and a browser fingerprint are stored only if you choose Allow. Clear conversation deletes it, including the archive copy. No name or email is required. AI suggestions may need correction. Privacy details.

Or browse all courses

A SMALL PROJECT. A REAL STEP FORWARD.

What will you
build first?

Choose a practical AI course.
Compare what you’ll build, what you’ll learn, and where to start.

Showing all 12 courses

Build Your First AI Agent — course cover illustration
Choose a useful task: Your task → Success criteria → Checkable result. Illustrative diagram.
Give the agent a job: Instructions → AI agent → First answer. Illustrative diagram.
Let it use a tool: Question → Approved tool → Evidence. Illustrative diagram.
Keep actions in bounds: Missing information → Ask or confirm → Act with permission. Illustrative diagram.
Test, then take it with you: New case → Check the result → Portable agent. Illustrative diagram.
1 / 6Illustrated preview

YOUR FIRST AGENT

Build Your First AI Agent

New to agents · Bring a work task

What you’ll buildAn agent that uses tools and asks before acting.

7 practical stages with an AI mentor

Is this for me?

Bring one work task and a way to judge the result. The mentor can handle the code.

What you’ll learn

Define a useful task, set action boundaries, and judge whether an agent’s output meets your criteria.

How you’ll check it

Try a held-out input, compare the result with your criteria, and explain what needs to change.

View course
Agent Skills: Beginner — course cover illustration
Choose a repeated task: Your method → Useful procedure → Clear result. Illustrative diagram.
Package your instructions: Instructions → Reusable skill → New input. Illustrative diagram.
Define where it belongs: Matching request → Scope check → Use the skill. Illustrative diagram.
Teach a difficult case: Tricky example → Improve instructions → Check again. Illustrative diagram.
Reuse on a new task: New material → Same procedure → Tested skill. Illustrative diagram.
1 / 6Illustrated preview

SKILLS · BEGINNER

Agent Skills: Beginner

No coding required · Bring a recurring task

What you’ll buildYour repeatable method, packaged as a reusable skill.

6 practical stages with an AI mentor

Is this for me?

No coding background required. Bring a recurring task and a few non-confidential examples.

What you’ll learn

Write reusable instructions, define when a skill applies, and check its output on a fresh case.

How you’ll check it

Use a reserved case to check scope and output, then download the package and verify a run outside the course.

View course
Agent Skills: Advanced — course cover illustration
Make activation visible: Two requests → Scope boundary → Right skill. Illustrative diagram.
Load only what is needed: Current task → Supporting guide → Relevant answer. Illustrative diagram.
Make skills work together: First skill → Handoff contract → Next skill. Illustrative diagram.
Compare on the same cases: No skill → Initial skill → Improved skill. Illustrative diagram.
Release with evidence: Source boundaries → Verified checkpoint → Tested release. Illustrative diagram.
1 / 6Illustrated preview

SKILLS · ADVANCED

Agent Skills: Advanced

Already have a basic skill

What you’ll buildAn improved skill, compared with the original on the same cases.

9 practical stages with an AI mentor

Is this for me?

Experience writing a basic skill. Advanced means skill-authoring experience, not a requirement to code.

What you’ll learn

Compare skill versions fairly, detect regressions, and decide whether an improvement is supported by evidence.

How you’ll check it

Compare the same cases without a skill, with the original, and with the revised version; test a new case before your release decision.

View course
Connect Your Agent with MCP — course cover illustration
Choose what to connect: Your task → Connection contract → Useful tool. Illustrative diagram.
Connect agent and server: AI agent → MCP connection → Tool server. Illustrative diagram.
Discover available tools: Tool list → Agent choice → Observed call. Illustrative diagram.
Make deliberate changes: Proposed update → Permission check → Confirmed action. Illustrative diagram.
Recover, then export: Broken connection → Honest recovery → Portable setup. Illustrative diagram.
1 / 6Illustrated preview

MCP

Connect Your Agent with MCP

Basic agent familiarity is useful

What you’ll buildAn agent connected to an external tool or data source.

8 practical stages with an AI mentor

Is this for me?

A task that needs an external tool or data source. First AI Agent is a useful starting point if agents are new to you.

What you’ll learn

Choose and inspect tool calls, set access boundaries, and diagnose a broken connection.

How you’ll check it

Change a data source and check the resulting tool calls and output against your integration criteria.

View course
Give Your Agent Memory — course cover illustration
Choose what should survive: Conversation → Remember deliberately → Saved fact. Illustrative diagram.
Retrieve the right memory: New session → Relevant memory → Useful context. Illustrative diagram.
Keep memories in scope: Task A → Memory boundary → Task B. Illustrative diagram.
Correct outdated facts: Old fact → Explicit correction → Current fact. Illustrative diagram.
Forget on purpose: Saved fact → Delete request → Verify removal. Illustrative diagram.
1 / 6Illustrated preview

AGENT MEMORY

Give Your Agent Memory

Basic agent familiarity is useful

What you’ll buildAn agent that remembers, updates and forgets deliberately.

8 practical stages with an AI mentor

Is this for me?

A task where information should survive between sessions. Basic familiarity with an agent is helpful.

What you’ll learn

Decide what to remember, verify retrieval, and check that corrections and deletions take effect.

How you’ll check it

Use fresh sessions and changed inputs to test memory boundaries and compare results with and without memory.

View course
From Meeting Notes to Verified Action Items — course cover illustration
Start with real meeting notes: Meeting notes → Your criteria → Useful actions. Illustrative diagram.
Tie each action to evidence: Source quote → Action item → Check the link. Illustrative diagram.
Keep missing facts visible: No deadline → Mark as unknown → Ask for clarity. Illustrative diagram.
Check owners and deadlines: Draft actions → Compare with notes → Verified items. Illustrative diagram.
Test on another meeting: New meeting → Same checks → Reusable assistant. Illustrative diagram.
1 / 6Illustrated preview

MEETING ACTIONS

From Meeting Notes to Verified Action Items

No coding required · Bring meeting notes

What you’ll buildMeeting actions with owners, deadlines and source evidence.

4 practical stages with an AI mentor

Is this for me?

Designed for managers who do not code. Bring anonymized meeting notes and your own criteria for useful action items.

What you’ll learn

Verify action items against meeting notes and recognize missing owners, deadlines, or cancelled commitments.

How you’ll check it

Test a new meeting, change one cancellation line, and explain which action items should change and why.

View course
Build Agents with the Claude Agent SDK — course cover illustration
Start with your learning goal: Your background → Project goal → Learning plan. Illustrative diagram.
Explore a tool-using agent: Instructions → Claude agent → Response. Illustrative diagram.
Connect tools to the task: User task → Tool request → Tool result. Illustrative diagram.
Coordinate specialized work: Main agent → Subagent and skills → Checked result. Illustrative diagram.
Build your capstone agent: Sessions and memory → Hooks and checks → Real-task agent. Illustrative diagram.
1 / 6Illustrated preview

CLAUDE AGENT SDK

Build Agents with the Claude Agent SDK

Choose your coding depth with the mentor

What you’ll buildA working agent built around the Claude Agent SDK.

12 practical stages with an AI mentor

Is this for me?

Choose a result-owner, strategist or coder track with the mentor. Expect SDK concepts and runnable experiments; coding depth follows your track.

What you’ll learn

Make and test decisions about tools, permissions, and agent behavior at your chosen coding depth.

How you’ll check it

Apply the SDK concepts in a capstone for your task; agree on the checks and your personal contribution with the mentor.

View course
Build a Voice Agent — course cover illustration
Choose a spoken job: Spoken request → Voice task contract → Useful reply. Illustrative diagram.
Take turns naturally: You speak → Agent listens → Agent replies. Illustrative diagram.
Change your mind mid-answer: Agent speaking → Interrupt → New direction. Illustrative diagram.
Confirm before acting: Tool result → Ask for confirmation → Test action. Illustrative diagram.
Recover without pretending: Missing result → Explain the issue → Browser voice app. Illustrative diagram.
1 / 6Illustrated preview

VOICE AGENTS

Build a Voice Agent

Microphone + browser · Live voice practice

What you’ll buildA voice agent that handles interruptions and confirmed actions.

8 practical stages with an AI mentor

Is this for me?

A microphone, a supported browser and a spoken task to test. Live practice sends audio to ElevenLabs.

What you’ll learn

Evaluate turn-taking, interruptions, tool-backed answers, and confirmation before an agent acts.

How you’ll check it

Try a held-out spoken task in a live session and compare the saved result with your criteria.

View course
Give Your Agent Knowledge — course cover illustration

AGENT KNOWLEDGE

Give Your Agent Knowledge

9 stages · Explore the program

What you’ll buildChoose a question worth grounding. Let the answer use your documents

9 practical stages with an AI mentor

View course

Explore the current published course catalog. Open a course page to review its program, starting point and next step. Your guest conversation stays on this website; it is not transferred to your app account.

THE LEARNING LOOP

Your own project.
Built, tested, understood.

Guidance when you need it.
Room to think for yourself.

  1. 01

    Start with your own task

    Choose a course around something you want to make useful. Define what a good result should do and what you’ll do yourself.

  2. 02

    Develop one working project

    Improve the same project stage by stage. Make a change, predict what will happen, and compare the result with your mentor’s help.

  3. 03

    Check what you can do yourself

    Test your project on new input. Judge the result, explain its limits, and identify what to change—using the checks in your course.

Your part in the work

You can focus on using and checking the result, designing instructions and behavior, or writing and debugging code. Agree on your contribution with the mentor for your course. Your coding experience and the work you want to do yourself are different questions.

In every approach, you define success and check what you create. A working project is one result; learning to use and evaluate it yourself is another.

BEFORE YOU START

A few good
questions.

Do I need to know how to code?

You can start with Agent Skills: Beginner, Meeting Actions or Build Your First AI Agent while the mentor handles code. You own the task and judge the results. Agent Skills: Advanced asks for basic skill-authoring experience, not programming expertise.

What should I check before choosing a course?

Review the starting point on its card, then select View course for the full program and access details. Bring a real task you can safely test.

Which tools will I need?

Guided web practice runs in the course workspace. Voice Agent also needs a microphone and sends live audio to ElevenLabs. Running an exported project outside the course can require your own runtime and provider account; the project README explains the setup.

How is the AI mentor involved?

The mentor helps explain concepts, build, and debug. Your contribution can focus on using and checking the result, designing its behavior, or writing code. Agree on that contribution for your course; you still judge the output and explain what happens on a new case.

What will I leave with?

It depends on the course. Examples include a reusable Claude skill, a no-code meeting agent, or a runnable MCP package. Look for the expected artifact in each course description.

Can I bring my own work problem?

Yes. The learning approach is built around useful work. Use anonymized examples and avoid sharing confidential or personal information that you do not have permission to use.

How much does it cost?

Sign in to try the introduction and first 2 learning phases free. No card required. Subscribe to continue. Current subscription plans are shown in the app; external provider costs are separate.

YOUR NEXT SKILL, PUT TO WORK.

What could you
make easier with AI?

Find a course to start

A practical first step starts with something you want to build.