Automations

Project Overview

Introduction

     I have ample experience in doing automation using various tools like n8n, Linux, Python, Playwright, databases, and different APIs. While I cannot share the exact JSON configurations or show specific workflows due to client confidentiality agreements (I take these agreements seriously), I can provide high-level overviews of successful automation projects I've delivered. These solutions have helped businesses save time, reduce errors, and focus on what truly matters.

Challenges

⏺ Client Confidentiality: Cannot share exact workflows or specific configurations due to client agreements.
⏺ Workflow Complexity: Managing complex automations requires careful design and modular thinking.
⏺ Reliability and Recovery: Ensuring workflows can recover from failures and maintain data integrity.
⏺ Maintainability: Creating systems that are easy to understand and modify over time.

Solution

     My approach to automation follows a microservices-inspired design philosophy. Instead of monolithic workflows, I split complex automations into modular, reusable components. Each workflow includes "savepoint" stages where data is persisted to databases for both logging purposes and the ability to rerun or restore workflows if failures occur. This approach ensures reliability, maintainability, and easier debugging. I also emphasize clear naming conventions and documentation through sticky notes within the workflows.

Key Features

✔ Modular Design: Microservices-inspired architecture for flexibility and reusability.
✔ Data Persistence: Savepoints enable workflow recovery and audit trails.
✔ Clear Documentation: Verbose variable names and node descriptions for maintainability.
✔ Error Resilience: Built-in recovery mechanisms to handle workflow failures gracefully.

Automation Examples

⚡ Daily Performance Report: Automated daily monitoring of agent performance reports sent to executives.
Workflow: Scheduled trigger → Playwright to extract data → Google Sheets → Playwright to screenshot dashboard → Send to Slack.
⚡ Report Writing: Automated creation of regular reports for clients.
Workflow: Trigger → Google Drive for files (images/data) → Python to insert into template → Google Drive upload → Email notifications.
⚡ Client Onboarding: Centralized internal tool for onboarding clients and users to platforms like Microsoft, CIPP, and Datto.
Workflow: n8n Form → HTTP for data validation → HTTP for posting payload → Logging to helpdesk.

Tips for People Reading This

1. Modular Design: Instead of having a big workflow, split it into multiple microservices-esque design for modularity.
2. Savepoints: Have a "savepoint" stage where you save data to a database for both logging and rerunning/restoring the workflow.
3. Use Set Nodes Liberally: You can never overuse "Set" nodes - they help clarify data transformations.
4. Clear Documentation: You'll forget, so have a verbose variable and node names/section description in sticky notes.