An online intensive by AIDOLAB

Take AI & Do build your own case in 16 workshops

A hands-on intensive for people who've already tried different AI tools and want to build AI into their work: you bring your own task and, step by step, assemble a working process, prototype, or automation around it.

  • 16 live 90-minute workshops
  • One case through the whole program
  • A group of up to 15 people, with feedback as you build
  • Every workshop moves your case forward
  • A prototype, automation, or AI agent at the finish line
Apply now See the program

Starts July 30 · 3 sessions a week · online · group of up to 15

where it all begins

What's a case

A case is a recurring, tedious, or long-postponed task that AI can simplify, speed up, or solve outright.

Example cases

Content pipeline

Turns ideas, notes, and research into posts, newsletters, articles, or scripts.

Knowledge base for your team's documents

Helps you quickly find answers in guidelines, decks, PDFs, notes, and working files.

Landing page or mini-service prototype

Lets you quickly test an idea for a product, course, event, or internal tool.

Learning assistant

Builds lesson plans, exercises, flashcards, quizzes, or a personal learning path.

Operations assistant

Helps you manage tasks, plan your week, process notes, and turn chaos into a clear action list.

Topic-monitoring agent

Tracks news, competitors, job openings, grants, events, or professional opportunities.

Every session moves your case forward. By session 5 the foundation is in place: a brief, tools, materials. By session 10 — a working mini-product. By session 15 — an automation or an agent-driven workflow. In session 16 you present your result to the group.

Not sure which task to pick? We'll talk it through with you before the course starts.

who it's for

Is this intensive right for you

This course is for people who've already dabbled with AI and want to build it into their work systematically — around a real task, not for a tour of tools.

A great fit if you

  • Are a manager or specialist who wants to work faster — and have something to show leadership
  • Are a team lead who wants to see where automation is genuinely useful and where it's just marketing
  • Are an entrepreneur or founder who wants to build a mini-product without hiring a developer
  • Have tried ChatGPT / Claude / Gemini, but never got past the "ask a question, get an answer" level
  • Can commit 4–6 hours a week and bring a real task of your own

Not a fit if you

  • Want an "overview of every AI tool" — we don't do tours of services
  • Expect a plug-and-play recipe that "just works"
  • Aren't ready to bring your own task — without a case running through it, the course doesn't work
  • Need turnkey code built entirely without you
program

Three modules. Sixteen workshops. Zero fluff.

The intensive is built like a staircase: foundation → prototype → automation. Each module ends with your case at a new level. Click a module to see the workshops inside.

M1 5 workshops · weeks 1–2

Foundation: from chat to a working tool

Learn to think before you prompt, frame tasks, verify facts, and work with your own materials. Assemble your starter stack and a first multimodal artifact.
  1. 1

    A map of the AI landscape, minus the overwhelm

    What actually matters about LLMs, agents, tokens, and context windows. Frame your case and assemble a minimal starter stack.

  2. 2

    Thinking before the prompt, and strong prompting

    The AI Task Brief as a "think → frame → verify" framework. Proven prompt formulas, practiced on your real tasks.

  3. 3

    Safety and AI literacy

    What's safe to upload to AI services, where you need fact-checking, and when it's better not to use AI unsupervised. Privacy settings and personal ground rules.

  4. 4

    Data, research, and your own knowledge bases

    Deep Research in Perplexity, ChatGPT, and Gemini. Working with PDFs and large documents. Building a mini knowledge base in NotebookLM.

  5. 5

    Multimodality

    Images, video, voice, and audio: what to pick and when. Build a multimodal artifact.

By the end of this module you'll have: a case card, an AI Task Brief, a personal safety checklist, a mini knowledge base, and your first multimodal artifact.

M2 5 workshops · weeks 2–4

Vibe-coding: from idea to working prototype

Build your first mini-product: a page, a form, an assistant, a knowledge base, a content pipeline, or an internal tool. All through AI services — no code written by hand.
  1. 6

    The vibe-coding map and token economics

    Which tool fits which task, and what it costs. Compare Lovable, Replit, Codex, and Claude Code. Pick your route.

  2. 7

    Your first prototype, no code

    Build a landing page, mini-service, or internal tool. Take the idea to a public link or a working draft.

  3. 8

    Agent-driven building without the fear

    How Codex and Claude Code work under the hood. Prepare your project for an agent: goal, structure, rules, and guardrails.

  4. 9

    Building a skill for your agent

    What Skills are and how they differ from a plain prompt. Turn a repeatable task into a working scenario.

  5. 10

    A complete mini-product

    Take one scenario to a verifiable result. Run it on a real example and document the limitations.

By the end of this module you'll have: a working prototype or mini-product with a public link or a working draft.

M3 5 workshops · weeks 4–5

Agentonomics and automation

Turn your mini-product into a controlled automation. Figure out where an agent genuinely belongs and where it's just marketing. Build a working loop.
  1. 11

    A deep dive into agentic systems

    Dissect real-world cases versus marketing promises. Pick the part of your process where an agent can genuinely help.

  2. 12

    What an agentic system is made of

    Tasks, roles, instructions, context, memory, permissions, logs, and human review. Using Hermes and OpenClaw as examples, build a minimal working loop.

  3. 13

    How it plays out in real cases

    Walk through examples from the course hosts: what they handed to an agent, what they kept under manual control, and why.

  4. 14

    Group practice, part 1

    Work on participants' scenarios, unpack the mistakes, and get to a first working result.

  5. 15

    Group practice, part 2

    Polish your automation into a final state you can show and explain.

By the end of this module you'll have: a working automation or a controlled process — the thing you'll present at the final session.

Finale session 16

Case showcase

Each participant shows the task they took on, what they built with AI, how it works, and where it can go next. The hosts and the group give feedback.

I want to build my case
what you'll walk away with

By the final showcase you'll have

Not lecture notes or "knowledge in general" — working artifacts you built with your own hands, for your own task.

  • An AI Task Brief and a prompt library

    a "think → frame → verify" framework tuned to your tasks

  • A mini knowledge base and a safety checklist

    your materials, organized so AI gives you accurate answers

  • A working prototype with a public link

    a landing page, form, assistant, or internal tool — no code

  • A controlled automation or agent

    a process where AI helps and you stay in control

  • A personal tool stack that fits you

    what to choose, what it costs, and how not to overpay for tokens

  • A presented case

    a real, working thing you can show at your company

course creators

Who runs the intensive

Sandra — founder of AIDOLAB and B5 Research

Sandra

Founder of AIDOLAB · B5 Research

Builds AI-powered products — from prototype to launch. Delivers corporate AI lectures for teams and executives. At B5 Research she's building a research practice where AI is a working tool, not a toy.

Nikolai — co-founder of AIDOLAB

Nikolai

Co-founder of AIDOLAB

Builds AI agents for business and studies the agentic economy. At AIDOLAB he owns the methodology and architecture of agentic solutions: how AI agents and coding assistants work, how to use them on real tasks, and how to design working systems on top of them. Develops educational programs at the intersection of AI, product thinking, and the practical application of technology.

Anastasia Tmur — AI literacy lead

Anastasia Tmur

AI Literacy Lead

Designs educational programs. Delivers lectures and hands-on sessions on AI tools for teachers, parents, and teams. Works at the intersection of education, technology, and practical AI.

Join the group
questions & answers

Frequently asked questions

Do I have enough experience? I'm not in IT.
The course is designed for people without a technical background. Vibe-coding and working with agents are set up so that no programming is required. What you do need is a real case and the willingness to spend 4–6 hours a week — 90 minutes per session, plus time on your own case.
What if I don't pick a case before the start?
That's exactly what the pre-work is for: a case-selection template, a video, and a questionnaire to help you decide. If you're still stuck, we'll choose one together in session 1. The course doesn't work without a case, so it's the first thing we spend time on.
What if I miss a session?
Recordings stay available through the end of the course plus one month after. At the next live session you can ask questions about what you missed. If you miss more than three sessions in a row, we'll help you catch up with a one-on-one review.
Is this for business or for personal work?
Both. The group will include managers, specialists, team leads, and entrepreneurs — everyone's case is different. The course isn't tied to one niche: the "your own case across 16 sessions" approach works for your day job and for your own business alike.
Do I need to buy paid subscriptions?
The base kit is the free tiers of ChatGPT, Claude, Perplexity, and NotebookLM. For Modules 2 and 3 we recommend subscriptions to Claude Code, Codex, or similar tools — around 25 dollars including taxes, though the amount can vary depending on your needs. Either way, we help you pick the best tools for the price so you don't overpay — and you'll learn not to burn tokens for nothing.
What kind of result will I have to show at work?
It depends on your case. It could be: an internal assistant for your team's routine tasks, a form with automated response handling, a content pipeline, a mini-site, or an agent-driven workflow for preparing reports. You'll present it at the final session — and then you can present it again, this time at your own company.
What language is the course in?
All sessions and materials are in Russian. The tools (Lovable, Claude Code, ChatGPT) are in English, but menu-level English is all you need.
Will the recordings be public?
Recordings are available to course participants only. Publicly, we share only promo episodes and clips from the final showcase — with participants' consent.
Can I get a refund?
Yes — up to and including session 3. If within the first three workshops you realize the course isn't for you, we'll refund the full price, no questions asked. After session 3 there are no refunds: by that point you've completed most of the foundation module and started building your case.
Apply via Telegram

Take AI & Do

Come with your own task. Leave with a working prototype, automation, or AI agent that you built yourself.

Starts July 30 · applications close once we reach 15 people

  1. Tap the button — our Telegram bot opens
  2. Answer 3 quick questions (name, contact, case)
  3. We reach out, discuss your case, and confirm your spot
Apply via Telegram

It's just an application — no strings attached. Takes a minute, right in Telegram.

Not a fit? Full refund

If within the first three sessions you realize the course isn't for you, we'll refund the full price, no questions asked. The refund window runs up to and including workshop 3.