Contact Information
Australia
- March 2, 2025
The Art and Science of Syncing with n8n: A Technical Deep-Dive
Syncing seems easy, right? Just grab data from System A, push it into System B, and you’re done. Bahahahaha! If only it were that simple. This document serves as both a practical guide to syncing strategies in n8n and a technical deep-dive into the specific challenges of synchronizing Microsoft Teams messages. It’s meant to be comprehensive, detailed, and reflect real-world implementation challenges beyond what you’ll find in typical integration guides.
Multi-Model, Multi-Platform AI Pipe in OpenWebUI
OpenWeb UI supports connections to OpenAI and any platform that supports the OpenAI API format (Deepseek, OpenRouter, etc). Google, Anthropic, Perplexity, and obviously n8n are not supported.
Previously I had written pipes to connect OWUI to these models, and n8n, but now I’ve combined all four into a single pipe.
This technical walkthrough explores the implementation of a unified pipe that connects OpenWebUI with Google’s Gemini models, Anthropic’s Claude models, Perplexity models, and N8N workflow automation.
The Open WebUI RAG Conundrum: Chunks vs. Full Documents
On Reddit, and elsewhere, a somewhat “hot” topic is using OWUI to manage a knowledge base / files and take advantage of OWUI’s built-in RAG (Retrieval Augmented Generation) functionalities. The thing is, sometimes, you’re not trying to retrieve snippets for context; you’re aiming for summarization, translation, file comparison, or brainstorming. I often see people struggling with system prompts or RAG prompts to get an LLM do process documents in ways that RAG simply doesn’t support. You can’t “chat to your PDF” and ask, “Take the grand totals section from my Excel file and re-write the summary in the financial report to reflect those numbers.” RAG isn’t built for this. It’s built to merely return chunks of text to the Agent that are hopefully semantically similar to the user’s request.
Code Smart in n8n: Programming Principles for Better Workflows
As a coder of 30+ years, I’ve learned that coding isn’t really about coding – it’s about logic, efficiency, and creativity under constraints. These fundamental principles are just as relevant in n8n as they are in traditional programming.
Sure, you can quickly learn n8n without any coding background. That’s one of its strengths! But understanding basic programming concepts can transform how you build workflows. Why? Because these time-tested principles help you create more robust n8n solutions.
Case Study – provider agnostic AI chat interface
A client wanted a “chatbot” to interface with all of the providers (Google, OpenAI, Perplexity, Anthropic, and image generation models) so they could easily compare
Creating a “Custom GPT” with Open Web UI
Why pay $20 per month to OpenAI when you can do it for “free” using OpenWeb UI.
People say “free” a lot in the AI space. It’s not really free, unless run OWUI locally, and you have the time and skills to set it up.
Upwork: A Freelancer’s Descent into Kafkaesque Absurdity
The gig economy. A promised land of flexibility, autonomy, and unlimited potential. Or so the marketing spiel goes. But what happens when the platforms designed to empower freelancers become instruments of arbitrary control, shrouded in vague policies and enforced by unyielding algorithms? This isn’t a hypothetical scenario. This is my reality on Upwork. And it’s a reality built on a foundation of fear, fueled by inconsistent enforcement, and ultimately designed to keep freelancers in a perpetual state of precarity.
Debugging OpenWebUI’s Hidden Image Processing Limits: A Technical Deep Dive
While integrating image generation with OpenWebUI using Replicate.com’s API, I discovered an undocumented processing limit. The system would fail when generating multiple 1440×1440 images but work fine with other aspect ratios. Through systematic testing, I uncovered that OpenWebUI has a hidden limit of approximately 4.7 megapixels for total output. This led to developing a dynamic scaling solution that automatically adjusts image dimensions based on the number of outputs while maintaining aspect ratios. The investigation revealed important insights about handling image processing limits in web applications.
Basic comparison between LLM providers
I’ve created this table of information to assist me in remembering the features and restrictions of each provider.
My issue with all AI providers / models
After testing multiple AI platforms, I realized they all had deal-breaking limitations—short conversation memory, prompt manipulation, censorship, rate limits, and unreliable file processing. Many chunk documents, making full-document interactions impossible, and frequently “hallucinate” answers instead of admitting uncertainty. Worse, AI providers prioritize storage efficiency over retaining context, making deep discussions frustrating. I needed an AI that could analyze entire documents, retain long chat histories, and offer real control—so I built my own system. Now, I have an AI that actually works the way I need it to.