Automation and Python

This service is for tasks where routine work should be handed to a system: leads, notifications, reports, calls, data exchange between services. For typical chains I deploy n8n on your server. When the logic is more complex, I write a Python service. For calling clients I build an AI assistant. All three can be combined in one project.

What you can order

Prices are per project. I give the exact scope and timeline after you describe the task, and we agree on them before work starts. Telegram bots are a separate service: Turnkey Telegram Bot Development.

Example from practice

An IT company that supports accounting software called clients with overdue payments by hand every month. An employee read out the same text, recorded the answers and called back those who did not pick up.

I built an automatic voice dialing system. It loads debtor data from a Google Sheet, builds the message text, voices it with speech synthesis, places the call through telephony, records the answer and writes the result back to the sheet. There are two modes: a fixed question-and-answer script and a free monologue from a text in the database. The client’s language is detected automatically, and dialing is retried up to 10 times at a set interval. The monthly calling of dozens of clients now runs without an employee, and the operator receives a table of results. Write-up: Robot collector: automatic debtor calls.

An example of n8n and Python together: a network of 150 Telegram channels for a travel agent. Scheduled publishing runs through n8n, while offer search, analysis and image generation are implemented in Python. Write-up: Automatic management of a Telegram channel network for a travel agent.

How the work goes

1. You describe the process: what triggers the automation, where the data comes from, where it should go. If needed, I help to formulate it.
2. I propose a solution: n8n, a Python service or a combination, the list of integrations, the timeline and the price.
3. I build it in stages and you check the intermediate result.
4. I launch it in Docker on your server, hand over access and documentation, and stay available for changes.

Related articles

n8n: choosing, installing, scaling

• n8n: Part 1 — Process automation without coding
• n8n: Part 2 — Cloud Version or Self-Hosted Server
• Installing n8n in Docker with HAProxy: A Clear Guide for Beginners
• n8n: Part 5 — Scaling: queue mode, workers and task runners
• n8n: Part 6 — Guardrails Node: Protecting LLM Workflows

When n8n is not enough

• n8n — not automation. It’s an orchestrator.
• n8n: a pretty wrapper that ate up two days

Python, queues and scraping

• RabbitMQ as a bridge between the external and internal networks
• How to Pretend to Be Human: Scraping Without Getting Blocked
• Data parsing in 2026: don’t run every page through an LLM!
• Why an LLM won’t replace a good parser. A case with auto parts
• Collecting data from websites and marketplaces: engineering data collection

Calls and voice robots

• How to set up queue dialing via SIP and Python
• Automated call transcript: from recording to a structured document

Frequently asked questions

How does n8n differ from Python development?

n8n suits typical chains: a lead, a CRM, a notification, a report. If you need complex logic, user roles, your own interface or heavy data processing, that is a job for Python. Often both are used: n8n runs the process and the heavy part is moved to Python, as in the channel network project.

Do I need my own server for n8n?

Yes, n8n is deployed on your server with Docker Compose. You need a VPS with at least 2 GB RAM; if you have no server, I can help choose one. There is no per-execution fee.

Can you improve a system that is already written?

Yes. I can join a project to improve or stabilise it: untangle legacy code, bring the architecture into shape, and add containers and monitoring.

What is needed for the AI call assistant?

A description of the business process and the call scenario, SIP telephony and a server for the system. I handle setup and testing. Client data is taken from a table in 1C, Excel, Google Sheets or Yandex Tables.

What about a Telegram bot?

Bots are a separate service: Turnkey Telegram Bot Development. The general approach to Python work is on the Python development page.

How to order

Message me on Telegram @BigMike_help: briefly describe the process you want to automate. I reply on the next business day. Other ways to reach me are on the contacts page.