A simple rule I use before automating any business process: write it down as steps a new employee could follow. If you can't, automation will only make the chaos faster.
Once the steps are clear, tools like n8n plus a few API calls can take over the boring parts: copying data between systems, sending follow-ups, building reports. The goal is fewer manual clicks, not more software.
Happy to look at a process you'd like to automate.
An AI agent fails halfway through a job. What should it remember when it tries again?
That question led me to build AgentTrace, a Python and FalkorDB project for keeping track of agent runs. It records task progress and decisions in a graph so you can inspect what happened and see where a run stopped. That’s especially useful when several agents share a workflow.
When you debug an agent run, what do you look for first: the failed step, the input that caused it, or the decision that sent it down that path?
Big week for me. On Thursday I'm being inducted into the engineering profession, and on Friday I graduate from the Federal University of Technology Akure after six years of studying Industrial and Production Engineering.
Industrial engineering taught me one thing above all: map the process before you try to improve it. Find where time is wasted, remove the steps nobody needs, then make the rest run reliably.
That's exactly what I do for businesses now, with AI automation in Make.com, n8n, Zapier and GoHighLevel. Same principles, different factory floor.
Grateful to everyone who helped me get here. I'm taking on new projects in automation, operations and video editing, so if your team needs a hand, my services are on my profile.
@Daniel Fajuyi Huge congratulations on the induction and graduation, Daniel. Six years of Industrial and Production Engineering is a serious milestone. That mindset of mapping the process and stripping out wasted steps before automating is exactly what separates clean n8n...