AI Code Automation

Programs I wrote with AI, running by themselves, saving hours.

Not a rented workflow tool. The logic lives in my own Python, built alongside an AI, mine to read and change. Three of them run every day, feeding live sheets and a desktop cockpit.

Python Monitor

A live desktop view of the machine and every running Python process, gauges and a table.

Trade Journal

Every closed trade written to a Google Sheet, one tab per strategy, live dashboards.

Control Cockpit

Strategies, executions, terminals and telemetry mirrored on one sheet, in real time.

The logic lives in my code

A drag and drop tool owns the plumbing. When it breaks, or the price changes, or a step is missing, you wait. Here the whole flow is a Python file I can open and rewrite the same afternoon.

A rented tool

Someone else owns it

  • The logic sits behind their interface.
  • A monthly bill, a hard ceiling.
  • You bend the job to fit the boxes.
My own code

I own every line

  • The flow is a file, read it top to bottom.
  • No ceiling, no per run cost.
  • Change it to fit the job exactly.
Ideawhat should run by itself
Built with AIcode written and reviewed
Runs on its own

Built with AI, plain Python

The same names the market asks for. A foundation model helps write it, then the program stands on its own: a loop, an API, a sheet, a JSON payload.

Language

Python and Node.

Wires

APIs, webhooks, JSON.

Models

LLMs and foundation models.

Output

Google Sheets, a desktop app.

automation.py
# my code, runs unattended
while True:
    state = read_live_system()
    data  = shape(state)      # to json

    write_local(data)         # my files
    push_to_sheet(data)      # api call

    sleep(60)                 # next pass

Three programs, running now

Not slides of what could be built. Real automations giving hours back, every day.

Python Monitor Trade Journal Control Cockpit