A live course with Saif Shraideh in the room, on-site or online. Each person sets up Claude Code on their own laptop, builds a personal AI operating system called a Home OS around their own work, and leaves running it.
For companies that send their people on courses to get better at their work.
Pick a question to run a session
A lot of people use AI like a search box. They ask, get something generic, fix it by hand, and lose the thread by tomorrow. They can tell they are getting a fraction of the tool and cannot see what they are missing.
The gap is the setup. Out of the box the tool knows nothing about the person or how they like things done, so it guesses and they do the rest.
Meanwhile the tools pile up. Each new one promises to fix the last, and managing them becomes its own job.
And everything a person works out lives inside a chat thread: hard to search, locked to one app, gone by the next session. The same ground gets covered twice.
A Home OS is a folder of plain text files on their own laptop, run by Claude Code inside VSCode. Nobody downloads a template. Each person builds it piece by piece and learns what every piece does: a file that says who they are and how they like things done, files that hold their projects and routines, rules that tell the AI how to behave, and small procedures it can run on its own. It lives on their machine and it is theirs. Because it is plain files, it also works with other AI tools, Codex for example.
The point is the structure. Once a person knows how the pieces connect, the AI does more than answer questions. It reads the right files at the start of a session, follows their rules without being reminded, makes the routine calls the way they would, and brings them the ones that need a human. They decide what it handles and what stays with them.
A month later the folder is bigger, because they know how to grow it. New project, new file. New rule, one line. Everyone in the room builds their own. Nobody copies Saif's.
Three people who took the course, and what a normal morning looks like now. The system is different for each of them, because each built their own.
Monday morning they open their folder and ask what moved last week. It has the notes from every call they logged, drafts the three follow-ups in their own tone, and flags the one account that went quiet. They read, fix two lines, send.
The call on the quiet account is still theirs.
A supplier slips a date. They drop the email into their folder and ask what it touches. It checks the delivery schedule they keep there, lists the two orders that move, and drafts the note to the customer in the wording they used last time.
They decide whether to tell the customer today or wait for the supplier's reply.
Before the weekly one-to-ones, they ask for the state of each person's work. It pulls from the project files they keep updated, shows what slipped and what shipped, and lists the things they said they would raise but have not yet. They walk into each meeting knowing what to ask.
What they say in the room is theirs.
People stop losing hours fixing what the AI gave them.
Each person's setup keeps improving on its own, so you are not buying training again next year.
What they build is in files, so a good way of working can be shown to a colleague.
Each setup learns how things are done at your company, so the output needs less fixing over time.
Your people see it as something the company did for them.
A call with Saif. He learns what your people do all day and what they use AI for now, then sends a proposal with the sessions and the price for your company. Before the first session each person needs a laptop and a Claude account on the Pro plan or above. The editor and Claude Code get installed together in the first session. No terminal experience needed.
Live, everyone building at their own laptop, Saif present. Each person installs Claude Code and builds a Home OS around their own work. Questions get answered as they come up, by someone who built his own the same way.
Each person leaves with a working system on their own laptop and keeps growing it themselves. They know how it works, so they extend and fix their own. That is why the course teaches the mechanism instead of handing out a template.
The course teaches Claude Code on the command line, inside VSCode, because of where the work ends up. Everything each person builds is a real file on their own disk. The editor and the terminal sit in one window, so a file gets fixed in the same place it lives.
The command line is also the base layer. The desktop app and the VSCode extension sit on top of it. Learn the base and the rest is familiar.
It goes past a conversation. From the terminal, the AI can run scripts and carry out steps on its own at set points, and the same folder works on any machine with Claude Code on it.
And the habit is the point. Someone at ease in a terminal with an AI next to them works differently from someone typing into a box.
I am Saif Shraideh, a mechanical design engineer from Amman with more than eight years in industry. I built Home OS for myself first, to run my own work and my own life, and I still run it every day. The course teaches the system as I use it, with nothing added for show.
I teach it one step at a time, checking each before the next. Each person builds their own setup in the room, and when they get stuck, I am there to look at their screen.
What you get from me is how the thing works, in plain English, so that your people can keep building it long after I have left.
Anyone whose day involves reading, writing, planning, or keeping track of things. It is for the people outside engineering as much as the engineers. If someone has never opened a terminal, that is expected.
None. The course starts from what to download. Nobody needs to know how to code.
A small group. Saif is in the room answering questions and looking at screens, and that sets the limit. The number is agreed with each company.
Per company, in the proposal that follows the call. There is no fixed price on this page because the group size and the number of sessions differ from one company to the next.
Each person needs VSCode and Claude Code installed on their laptop, which usually means admin rights or a word with IT beforehand.
Either. Saif comes to your office, or the cohort runs online, live with Saif either way.
Before the course, you tell Saif what your people do all day. The examples in the room are their own work, and each person builds their system around their own tasks, not a demo.
That is the normal case. Everyone builds their own system on their own laptop, so someone who is quick keeps going while someone who is stuck gets help. Nobody waits for the room to catch up.
The system each person builds holds their own projects, notes, and routines, so it is only ever about their work. Saif learns what your people do before the course so the examples fit.
Each person's setup is a folder of files on their own machine. Nothing is uploaded to Saif and he keeps no copies. What each person types into the AI goes to Anthropic under the terms of the Claude account they use, and whether it is used for training depends on the plan, so check that with whoever owns the account.
Each person leaves with a working system and keeps growing it themselves. It is theirs, on their laptop, and it does not stop working when the course ends.
Ask each person what their system does for them now. Ask again a month later. If the answer has grown, it worked.