Skip to the main content

People should have
a hand in the work.

Pattern Works builds visual AI systems around the people who know the work. Our preferred tools are n8n for agents and workflows, Airtable for review interfaces and GitHub for shared guidance and skills.

Hands sketching ideas on paper beside orange and pink sticky notes.
Illustrative photography: FORTYTWO / Unsplash

The experience behind Pattern Works

Pattern Works was founded by Julius Crutchley, an AI workflow and automation engineer with experience building systems at an agency serving major entertainment brands. His work spans content planning, publishing automation, video quality assessment and structured data ingestion.

His entrepreneurial achievements include an Innovate UK Young Innovators Award, winning StartUp Sussex and receiving a Santander Elevate Grant.

That entrepreneurial background shapes how Pattern Works approaches AI automation. The starting point is a problem worth solving: who experiences it, what makes it difficult and whether a proposed system would make their work meaningfully better. Prototyping and testing help establish that before committing to a larger build.

Commercial judgement continues through the engineering. Model costs, development time, maintenance and the effort needed to check an output all affect whether an automation is worthwhile. Pattern Works brings those considerations together with hands-on technical experience, building around the people who understand the work and will depend on the result.

A direct working relationship

Whether you work on your own, run a small team or improve processes inside a larger business, you work directly with the person designing and building your system.

Your expertise has a place

Your know-how shapes the agent’s rules, reference material and skills. We work through examples with you to make that knowledge usable, while keeping your judgement involved in the results and decisions that matter.

Visual work. Reviewable results.

Visual n8n workflows make the process available to follow. Airtable interfaces can put results in front of you to check and correct. A shared GitHub library records changes to guidance and skills. We also use Supabase and adapt to existing tools where they fit the work.

See the approach in practice

Living Article is our own development project. It provides concrete examples of research, writing, source inspection and engineering investigation. The Work page explains what those examples demonstrate and where further review is needed.

Explore the approach, or tell us about a piece of work you would like to improve.

See how we work