← Selected workCASE STUDY 02 / 03

Workflow automation · Geneseeq Technology

Giving time back to operations.

AI-assisted data extraction and a human review step made a high-volume ordering workflow substantially faster.

MY ROLE

Workflow design, AI implementation & operational delivery

PERIOD

Geneseeq technology leadership

~60,000Annual orders in the workflow
5 minPrevious entry time per order
30 secEntry time with AI assistance

01 / The context

The operating challenge.

Processing patient orders created repeated manual data-entry work. At approximately 60,000 orders per year, even small improvements in each transaction could materially change the operating workload.

The workflow needed to preserve a clear role for the people responsible for reviewing extracted information. Faster processing had to fit the existing operating process.

02 / My contribution

What I was accountable for.

I implemented locally hosted AI-assisted processing within the ordering workflow, connecting extraction with user review and the systems used by operating teams.

The work combined workflow design, enterprise integration, and production ownership. The objective was a usable change in daily operations, with human review retained for the extracted data.

03 / Key decisions

The choices behind the work.

Improve a recurring operating task

The use case focused on a repeated, high-volume activity with a clear before-and-after measure: time spent entering each order.

Keep human review in the workflow

Users reviewed the extracted information. AI assistance reduced entry effort while retaining an explicit review step.

Fit the surrounding systems

The capability operated within the ordering process and its enterprise context, where data quality and accountable ownership remained part of delivery.

How the workflow connects

  1. 01Receive

    Order information enters the workflow

  2. 02Extract

    Locally hosted AI assists data entry

  3. 03Review

    A user checks the extracted data

  4. 04Continue

    The order moves through operations

A simplified overview of the operating workflow.

04 / What changed

Results in the operating environment.

The workflow covered approximately 60,000 annual orders. Entry time decreased from five minutes to approximately 30 seconds per order.

The before-and-after times describe the entry activity. They are not measures of the entire order-to-report process.

05 / The lesson

A well-chosen operating workflow can make the value of AI tangible. The integration and review experience are central to turning a model capability into sustained everyday use.

Figures are approximate where indicated and reflect the period described.