AI & intelligent workflows

AI should help people move forward.

My interest in AI is not about adding intelligence to a product because the technology is available. It is about finding the moments where people spend time doing work that machines can assist with.

01Where AI becomes useful

Finding information
Extracting information
Organising information
Generating a starting point
Checking the result

In document-heavy professional workflows, this can reduce repetitive effort while preserving human judgement.

02Imagility examples

OCR & intelligent document extraction

Extract relevant information from documents, reduce repetitive data entry and bring information into the appropriate workflow.

AI-assisted RFE Response Builder

Attorneys responding to Requests for Evidence may need to review large amounts of information and construct detailed responses. The AI experience supported starting points, contextual prompts, suggested content, response assistance and user-controlled insertion.

Document
OCR / Extraction
Structured information
AI assistance
Expert review
Expert decision

03My AI design principles

  • Start with the workflow, not the AI capability.
  • Use AI where it removes repetitive effort.
  • Make the AI contribution visible and understandable.
  • Keep the expert in control of consequential decisions.
  • Design review and correction as first-class interactions.
  • Avoid creating a separate AI experience when AI can fit naturally into the existing workflow.

"AI is most useful when it disappears into the workflow and simply helps people move forward."