Arman Afzar

AI Copilots & Agents

Automation without a baseline: why AI projects can’t show results

If you don’t know how long a process takes today, you’ll never be able to say how much automation helped.

The Arman Afzar team2 min read

Many automation projects hit a simple question after a few months — “how much did we save?” — and have no clear answer. Usually it’s because nobody measured the current state before starting.

What a baseline measures

Three numbers are enough: how often the task happens each week, how long each one takes, and what share needs rework or re-routing. Those three numbers become the pilot’s success metric.

Process first, then tools

Automating a messy process just makes the mess faster. Before building an n8n workflow or an agent, map the current path, remove unnecessary steps and mark the points that need a human decision.

An example

Take expense invoice entry. The baseline says how many minutes each invoice takes and how many are entered wrongly. A bot like Sabtaa extracts and records ten invoice fields from a photo; with a baseline you can say exactly what changed.

Related articles

  1. 01AI copilots fail without a knowledge spineMost enterprise copilots don’t fail because of the model — they fail because knowledge lives in ten different places. What a knowledge spine is, and how to build one.
  2. 02An LLM on your own servers: where to startWhy organisations run language models on-premise, and what a good pilot looks like.
  3. 03A support copilot pilot: a realistic planFour stages that take a support copilot from demo to an everyday team tool.

Tell us your project in one sentence — we’ll find the right facet together.

Start the conversation