DEV Community: Cerewro Team

What an AI bot can do with its own Linux computer for €39/month

Cerewro Team — Wed, 28 Oct 2026 19:23:56 +0000

Most AI assistants stop at text. You ask, they answer, and then you copy the code, run it, fix the error, open the spreadsheet, and save the file.

We built POSITROM differently. In Bot Mode (€39/month) every customer gets an AI bot with its own private Linux computer in the cloud: a real desktop, a terminal and a file system. The bot doesn't just tell you what to do. It does it on its machine and hands you the result.

Here are five real tasks it handles.

1. Turn a folder of PDF invoices into a clean spreadsheet

Upload 40 invoices and ask: "Extract date, supplier, tax ID, base amount and VAT into Excel."
The bot opens each PDF on its computer, parses it, builds the sheet, checks the totals and gives you an .xlsx to download.

2. Write a script, run it, and fix it until it works

"Write a Python script that renames my photos by date taken."
A normal chatbot gives you code and wishes you luck. The bot writes the script, runs it in its own terminal, reads the traceback if something fails, fixes it and only then returns the working script plus the output.

3. Collect public data and deliver a CSV

"Get the list of public holidays for 2027 in these five countries and put it in one table."
The bot opens a browser on its desktop, gathers the data, cleans it and saves a CSV/Excel file in its workspace.

4. Produce documents end to end

"Create a 10-question exam on photosynthesis with an answer key, in Word, plus 8 slides for the class."
The bot writes the content, formats it into .docx and .pptx files on its machine and leaves them ready to download.

5. Prototype a small web page or API

"Build a landing page for my bakery with a contact form."
The bot creates the project files, serves them locally on its computer to check that they work, and hands you the folder.

It also works from your phone

The same computer is available from the Android app, with voice and notifications when a task is done.

Watch it work (every video was recorded on the bot's own computer)

Everything shown in these videos was done by the POSITROM bot itself, on its own Linux computer, no human editing the files.

30-second demo: https://positrom.com/media/positrom_bot_demo_en.mp4
For accountants/agencies: https://positrom.com/media/sectores/gestorias_en.mp4
For academies and teachers: https://positrom.com/media/sectores/academias_en.mp4
For developers: https://positrom.com/media/sectores/programadores_en.mp4
For content creators: https://positrom.com/media/sectores/creadores_en.mp4

Pricing

Bot Mode: €39/month (a bargain: less than a single hour of freelance work): AI bot + its own private Linux computer (desktop, terminal, files), 35M tokens.
Chat Mode: €12/month: if you only need the chat (voice, PDF/Word/Excel/PowerPoint downloads).
300,000 free tokens to try it, no card needed. Cancel anytime.

👉 https://positrom.com/?lang=en

What task would you give a bot that has its own computer? Tell me in the comments and I'll run it and post the result.

How an AI agent runs its own Linux desktop and terminal to finish tasks

Cerewro Team — Mon, 26 Oct 2026 21:01:06 +0000

Most assistants stop at text. They answer, you copy, you paste, you do the work. We wanted the opposite: an agent that gets its own computer and does the task itself. This is a walkthrough of how that works in practice: how an agent operates a private Linux desktop and terminal, and how it proves that it actually finished.

Why text-only wasn't enough

Ask a chat assistant for "a Word report from these three websites" and you get a wall of text. The real task has more parts: open the pages, extract what matters, write a file in the right format, check that the file exists and opens, then tell you where it is. Those are actions on a machine, not sentences.

So the agent gets a machine. One private Linux computer per user, with a graphical desktop, a terminal and a file manager.

The workspace: four tabs

The user watches the agent through four views of the same machine:

Agent: the conversation and the step-by-step log.
Terminal: the shell the agent types into.
Files: the agent's working folder, where results land.
Computer: the live graphical desktop, in the browser.

The Computer tab matters more than we expected. A log tells you what the agent claims it did. A live desktop lets you see it. Trust comes much faster from the second.

The loop: plan, act, verify, report

Every task runs through the same four phases.

1. Plan. The agent splits the request into small steps and picks the tool for each one: browser for reading pages, shell for scripts, a document library for Word/Excel/PDF output.

2. Act. It executes one step at a time on the desktop or in the terminal. Each step has a hard cap, and the whole run has a step limit (25 steps per run) so a confused agent can't spin forever.

3. Verify. This is the phase that separates an agent from a chatbot. After writing a file, the agent doesn't say "done". It checks:

ls -la ~/results/report.docx          # does it exist, is the size plausible?
python3 -c "import docx; d = docx.Document('report.docx'); print(len(d.paragraphs))"

If the file is missing, empty or unreadable, the agent goes back to the act phase instead of reporting success.

4. Report. The final message says what was produced, where it is, and anything it could not do. The file is in the Files tab, ready to download.

Long outputs: write in chunks

Our first version tried to write a ten-chapter document in one go. Long generations got cut off in the middle, and the file was silently truncated. The fix was boring and effective: write the document in parts, appending each chapter to the file, then run the verify step on the whole thing and retry any part that is missing. Chunked writing plus verification turned a flaky feature into a dependable one.

Isolation as a principle

Each user's machine is theirs alone: their own disk space, their own processes, their own files. A user only ever sees their own machine, and the agent can only touch what is inside it. We treat that as a product rule, not an optimization. If an agent can run arbitrary commands, the blast radius of a mistake must be one user's sandbox.

Mobile

Not every task starts at a desk. There is an Android app where you give orders by text, voice or attachment and then check the result later, from the same files the desktop sees.

What we learned

Verify outputs, don't trust them. Checking the file is cheaper than apologizing for a broken one.
Chunk anything long, and retry on failure.
Show the real desktop. Visibility is the cheapest trust feature there is.
Keep the chat input pinned and the header collapsible; the conversation needs the screen.

Try it and tell us what breaks

This is the idea behind POSITROM AI: a bot that gets its own private Linux computer and finishes tasks there while you watch. If you want to try it, it's at positrom.com. I'm on the team, and I'd like to know which task you'd hand it first and what would stop you from trusting it.

What we learned taking AI from a free desktop app to small businesses

Cerewro Team — Thu, 15 Oct 2026 16:52:12 +0000

We're the team behind Cerewro, a free AI assistant for Windows. Building it, one question kept coming back from users: "Can my company use this for real work?"

Three lessons shaped our answer, AITarea:

Businesses don't buy AI, they buy time back. Nobody asked for features; they asked for fewer hours spent on proposals, reports and repetitive admin.
Data has to stay home. For many SMBs the blocker isn't price, it's sending client files to someone else's cloud. AITarea runs on the company's own PCs.
Adoption needs a human. Tools alone don't change habits, so we pair a simple plan with a free assessment and hands-on setup.

The result is a simple offer: a business plan with a 14-day free trial, and custom setup for teams that want AI in their own processes.

If you work with SMBs, we'd love your take: what's the first process you'd hand over?

https://aitarea.com/?utm_source=devto&utm_medium=articulo&utm_campaign=lanzamiento_aitarea

Disclosure: we build AITarea.

Building a free Windows AI assistant: one window for cloud and local models

Cerewro Team — Wed, 14 Oct 2026 22:17:36 +0000

We're a small team in Granada (Spain) and we've been building Cerewro, a Windows 10/11 desktop app that lets you talk to several AI models from a single window.

Why we built it. We kept juggling browser tabs for different assistants and wanted one place on the desktop, with the option to run a model locally when you don't want data leaving your PC.

How it works (high level). You choose a provider per conversation: a cloud model with your own API key, or a local model with no key. Keys are stored on your machine. Switching provider doesn't change how you work.

What we learned. (1) Asking for less at sign-up matters: we cut it down to email + password. (2) Being upfront about cloud costs avoids confusion: the app is free, the AI provider bills its own API usage.

Disclosure: I'm on the Cerewro team. It's completely free (no plans, no card). Feedback on what's missing or confusing is very welcome: https://cerewro.com