OctaLoom · Workshop

Claude Code workshops for organizations

A program built around your organization. It starts with a deep discovery of your tools, teams and workflows, and continues into two sessions where people work in Claude Code on their own real tasks and leave with a process that runs.

A deep discovery for every organization before anything is built
Content shaped around your tools and workflows
2 sessions · 4 hours each
Everything in the Claude desktop app

Trusted by

Organizations and communities that already ran the workshop, alongside dozens of individual participants. Their reviews are below ↓

The gap

The gap in most organizations

Everyone has sat through a talk about AI. Far fewer can open the tool on a Sunday morning and do something with it that touches their actual work. That gap is what this workshop closes.

They know the name, nothing else

They can say Claude, and still have no idea where it lives, what it can see, or what it can actually do to their files.

One try, then they quit

They get a mediocre result, conclude the tool is weak, and go back to working by hand, when almost always it is the request that needed the work.

No method, no continuity

Everyone invents their own way, and whatever worked for one person stays with that person instead of becoming how the team works.

Discovery

Every program starts with discovery

Before I build a single session, I learn the organization: a call with whoever is driving this, a short questionnaire for the participants, and a call with IT where it is needed. The examples, the exercises and the integrations of both sessions come out of that, so every organization gets a program that fits the way it already works.

  1. 1
    Tools and systemsWhat the team opens every morning (Microsoft 365 with Outlook and OneDrive, Google Workspace, Slack, Notion, a CRM and the rest), and which of those can connect to Claude.
  2. 2
    Technical orientationWho already works with AI, who is touching it for the first time, and the right pace for each group.
  3. 3
    WorkflowsThe tasks that repeat every week and eat the most hours. One of them becomes what we build in the second session.
  4. 4
    Who is in the roomWhich teams and roles are coming, and what each team needs to take back to its own work.
  5. 5
    Security and sensitive dataWhat must not leave the building, your IT policy, and which permissions the tool gets.
  6. 6
    What counts as successWhat management wants to see one month after the workshop.

Structure

Two sessions, both on your material

The fundamentals of Claude Code are the same everywhere, but the examples, the exercises and the integrations come from the discovery. The first session is about working the tool on tasks from each participant's world. The second is about building the processes we picked together.

01

Claude Code fundamentals

Session one · 4 hours

The goal: everyone in the room finishes knowing how to run the tool on their own. Four hours, because half the time goes to hands-on work on the machine.

  • Making sure everything works. The app is installed in advance using the guide I send, and in the first minutes I check that everyone has it open, signed in and ready.
  • Where this lives. One app with three ways to work, and when you move to the Code tab.
  • One folder, and only that folder. Claude Code sees only the folder you point it at. That is both the protection and the limit.
  • The loop. It reads, plans, acts, checks itself, and repeats until the result matches the request.
  • Permissions and approvals. When it asks, when it acts on its own, and what it must never touch.
  • How to phrase a request. A vague request costs more time and more money. We practice phrasing and look at what comes back.
  • Plan mode. Asking it to show a plan before it touches anything.
  • Context and memory. What it remembers, what gets dropped, and when to start a fresh conversation.
  • The file that travels with you. CLAUDE.md as a standing handover for a worker who shows up every morning with no memory.
  • Commands worth knowing. The slash commands that save the most time day to day.
  • Guided practice. Every participant runs two or three real tasks from their own world while I move between them.
You leave withThe app installed and working, a working folder of their own, and a first CLAUDE.md they wrote themselves.
02

Deep practice, on your own tasks

Session two · 4 hours

The goal: take a process that repeats every week and turn it into something the tool runs. This is where it moves from talking about capabilities to building.

  • Connecting to what you already have. Claude connects to the tools we mapped in the discovery, so it works on the team's real material.
  • Skills. A process you have explained three times becomes a standing skill the tool runs on its own.
  • Building a real skill. Each team takes the process chosen in the discovery and builds it during the session.
  • Dedicated helpers. A heavy task gets sent off to the side and comes back summarized.
  • Automations. Tasks that run by themselves on a schedule or on an event.
  • Safety and control. What to check before installing anything external, and which actions always require approval.
  • How it carries forward. What gets saved, how to hand it to a colleague, and how it becomes the team standard.
You leave withAt least one organizational workflow that is built, tested and running, saved in a folder and ready to hand to the rest of the team.

Who runs it

A bit about me

Hanita Yudovski

Hanita Yudovski

OctaLoom · B2B Marketing & AI Agents

I build AI agent systems for B2B marketing, and I run LinkedIn for companies: the CEO's profile, the company page, and employee ambassador programs. The tools I teach are the ones I run on my own business every day, so every example in the workshop comes from a system that is built and running here.

I moved to Claude Code when I hit the ceiling of what chat could do. Going from asking a model a question to a tool that does the work inside my own folder is exactly what this workshop teaches, and I have already taken more than 100 participants through that move, in organizations, in communities and in open workshops.

100+ participants
Workshops in organizations and communities
LinkedIn-First B2B

What they said

Dozens of individual participants have been through this

Alongside the organizations, I have run this workshop for dozens of individuals: marketing managers, content creators, video editors and musicians. Here is some of what they wrote at the end, published with their consent.

Whatever your level with AI, you need to listen to Hanita. She got everything across clearly to everyone, no matter what they walked in knowing, with patience, charm, humor and real accessibility. In a world of gurus selling for the sake of selling, there is actual value here.
Inbal ZoliContent creator
The workshop finally made me take a step forward and stop postponing my first real meeting with Claude. Hanita is available, attentive and extremely agile, and it shows in how she teaches. She identifies the need, adapts the content to the room, and adds the extra that makes it far more practical.
Pnina KoganGlobal marketing manager
Highly recommended. She organized exactly the points I needed in order to start, and turned something that felt complicated into something accessible. The atmosphere was easy and safe for asking questions, even remotely. I left with much more confidence and curiosity to keep going.
Alina ZakharovVideo editor
I left every session lit up with my brain running at 200. Even coming in with prior knowledge, I still walked away with more. I am already getting day-to-day value out of what I learned.
Or ‘OBD’ Ben DavidVisual storyteller
I came in with no Claude experience at all and learned a huge amount. I now have real motivation to work out how Claude serves me and makes me a better marketing manager. The amount I picked up genuinely surprised me.
Noga RubanenkoNonprofit marketing manager
You were excellent, with a lot of patience and a lot of knowledge that you filtered carefully as you went. The hands-on part was essential for me. It was very cool and opened up a completely new world.
Guy NaorMusician and engineering student
I did not expect to learn this much in a single session. Hanita explains everything at eye level, clearly and practically, and gives a lot of value.
Hila DaganMarketing manager

Scope

What this is, and what it is not

Yes

  • Every participant with a laptop open, working as we go.
  • A program built from the discovery of your organization.
  • Examples taken from your content and your work.
  • Follow-up material: an installation guide, the deck, and a practice sheet that stay with you.
  • A pace that works for people who have never touched this.

No

  • An inspirational talk about the future of AI.
  • A tour of twenty tools without going deep on any of them.
  • An off-the-shelf program that runs identically everywhere.
  • A coding workshop. No technical background needed and no code is written.
  • A session that ends without something working coming out of it.

Logistics

What it takes to run this

FormatOn site only, so that during the hands-on parts I can move between participants.
Group sizeOne instructor per five participants, so everyone gets attention during practice. Up to five I run the workshop alone, and at ten I bring an additional instructor. Each additional instructor is priced separately.
Gap between sessionsOne to two weeks. Long enough to try things alone, short enough not to forget.
What participants bringA laptop with the Claude desktop app installed and an active account. An installation guide goes out a few days before the first session.
Security and dataWhat can be connected and what stays out is decided in the discovery, together with whoever owns IT on your side.
What the organization arrangesLicensing for the tool and IT permissions. I walk that through with whoever owns it on your side.

FAQ

Questions that keep coming up

Do I need to know how to code to attend?

The name is misleading. Claude Code works on the folder you point it at and the files inside it: the documents, spreadsheets and decks your team opens every morning. In the first session everyone in the room learns to run the tool on their own tasks, and in the second we build the workflows chosen during discovery.

What is Claude Code?

An Anthropic tool that lives in the Claude desktop app, under the Code tab. It sees one folder you choose, then reads → plans → executes → checks itself → and repeats until the result matches the request. Permissions are set in advance: when it asks, when it acts on its own, and what it must never touch.

How long is the workshop and how is it structured?

Two in-person sessions, four hours each, one to two weeks apart. The first covers the fundamentals, the second builds the workflows chosen during discovery. The gap is long enough for people to experiment on their own, short enough that they do not forget.

Why is there a discovery phase before the workshop?

Because a general AI talk has already been through your organization, and few people can open the tool on a Monday morning and do something that relates to their actual work. In discovery I learn what the team opens every morning, which recurring tasks eat the most hours, who already works with AI and who is touching it for the first time, and what management wants to see a month later. The examples, exercises and integrations for both sessions are built from that.

What do participants need to bring?

A laptop with the Claude desktop app installed and an active account. The install guide goes out a few days before the first session, and in the first few minutes I check that everyone has the tool open and connected.

What happens with sensitive data?

Claude Code sees only the folder you selected, which is both the protection and the limitation. What can be connected and what stays out is decided during discovery, together with whoever owns IT on your side.

Who handles licensing and permissions?

The organization. Tool licensing and IT permissions sit on your side, and I walk the process through with whoever owns it. Start early, it is usually the slow part.

How many people can attend?

One instructor per five participants. Up to five I run the workshop alone, and at ten I bring an additional instructor, so everyone gets attention during practice. The workshop is in person only, for exactly that reason.

How is this different from a regular AI workshop?

A regular AI workshop shows you tools and you leave with a list. Here participants work on their own real tasks and on the organization's own files, and leave with a workflow that runs. That is also why discovery comes first: without knowing what the team opens every morning, every example in the room is somebody else's example.

Is this vibe coding?

Vibe coding means describing what you want in plain language and letting the tool build it. That is what happens in the workshop, except the output is not an app: it is the process you run every week, the report you used to assemble by hand, the connection between two systems that do not talk. Most of the first session is about exactly that, how to phrase a request so what comes back is what you wanted.

How is this different from the ChatGPT we already have?

ChatGPT answers you in a window, and you copy that answer over to where the work actually lives. Claude Code sees the folder you selected, opens the files inside it, works on them and hands back a result. What changes is who does the work between the answer and the file.

Ready to run this with your team?

A first discovery call, and after it a proposal built on your teams, your tools and your workflows.