Sylog
About

About Sylog

Sylog is a cloud coding agent you use in a web chat. Every project gets its own cloud computer, a full Ubuntu desktop, and the agent does its work there while you watch.

Who it is for

Sylog is meant for developers and small teams who hand work to a coding agent and want to see how it goes. It has no customers yet: it is in early access and does not admit the public.

The problem

Coding agents mostly work where you cannot see them. You hand over a task, wait, and get back a result or an error. When something goes wrong there is little to inspect and nothing to take over.

Sylog puts the agent on a real computer that belongs to your project and that you can open at any moment.

How it works

  • Each project has its own cloud computer with a desktop, a terminal, a browser, files and internet access. It sleeps when idle and wakes with your files and installed programs in place.
  • You describe the work in the chat. The agent writes and runs code in the terminal, works with the project's files and fetches web pages, and shows every step in the chat as it happens. Driving the desktop and its browser is on the roadmap.
  • You can watch the computer's screen live, take over its mouse and keyboard, and hand it back.
  • The models run on Google Vertex AI. You choose the model and an effort setting.

You write to the agent in the chat. The agent runs commands on your project's own cloud computer. You watch that computer's screen, terminal and files live, and you can take over its mouse and keyboard at any time.

A diagram of how Sylog works, not a screenshot of the product. Real frames from a run follow below.

The product, as it is today

A Sylog chat: the request to write fib.py and its unittest tests, then the agent's steps on the project computer -- a bash command, file writes and reads, and a bash run of the tests -- each marked done
A request in the chat and every step the agent takes on the project's own cloud computer, as it happens.
The end of the same Sylog chat: the agent's last steps and its answer, with the unittest output 'Ran 5 tests ... OK' and the first 15 Fibonacci numbers
The agent ran the tests it wrote on that computer and reports what it verified.

Where Sylog is today

Sylog is in early access and does not admit the public yet. Every change is deployed to its development environment, which is live on Google Cloud, and only the operator's test accounts can sign in there. There the agent already writes code and runs commands and tests on a project's own cloud computer, as the screenshots on this page show. Requests for early access join a waiting list until sign-up opens at launch.

Payments are not open yet. The plans and their prices are on this site; running work needs a plan, and plans open at launch.

What it runs on. Sylog runs on Google Cloud: its services run on Cloud Run, its database is Cloud SQL for PostgreSQL, and its agent calls Gemini models on Vertex AI.

Roadmap

The order we are building in. There are no dates here: each step ships when it works.

  1. First

    • Computer reliability: every project computer sleeps, wakes and recovers within its bounds, proven by automated tests on real machines.
    • Effort settings that work the same way on every listed model, and switching the model mid-chat.
    • Attachments for every model: images, PDFs, office files, audio and video.

  2. Then

    • Every model sees and drives the desktop and the browser.
    • Organizations: invite people with a role, and share or restrict projects and chats.
    • Safety: secret values kept out of the model's view, a card before consequential actions, and an always-on watch over what the agent does.
    • Each chat chooses its computer's cores and memory.

  3. Later

    • Partner and open models on Google Vertex AI beside Gemini.
    • Workflows: several agents working on one large task, on one card you can stop.
    • Automations that run on a schedule or on a trigger.
    • Memory across chats, and extensions such as skills and MCP servers.

Company

Sylog is built and operated by its sole founder, Maksim Mikhnev, as an individual.

The first commit in the project's repository is dated 14 May 2026. Sylog is in early access: sign-up is not open to the public yet, and the product runs every day in the founder's own environment.

Outside this site, the founder's public profile is on GitHub.

How to reach us and ask for early access: Contact.

Write to us at contact@sylog.app