The manual redaction step, removed

Automated text extraction and redaction for a USA-based credit servicing company, built on ASP.NET Core and Angular with RabbitMQ carrying the volume.

Talk to an engineer Data engineering & ML

Consumer credit Client

Data engineering & ML Practice

April 2025 Published

01 — The challenge

Where it started.

Sensitive customer documents were read and redacted by hand before they could move through the workflow. Throughput was capped by headcount, and errors surfaced downstream where they were expensive.

There was also no reliable control over who could see which document at which stage of the process.

02 — What we built

The work itself.

OCR across incoming PDFs. Text extracted automatically from documents that previously had to be read in order to be processed.

Sensitive fields removed in the pipeline. Redaction applied as a workflow step, so the protected version is produced rather than prepared.

RabbitMQ queueing. Document volume absorbed by a queue rather than by the request path, so a spike is a delay rather than a failure.

Role-based user management. Access and workflow state bound together, which is what makes the compliance position defensible.

03 — Results

What the client has now.

Manual effort and processing time reduced substantially. High document volumes handled through queueing rather than headcount. Role-based access strengthened compliance and collaboration across teams.

The rest of this study

What we built, what it produced, and the stack behind it.

We share the whole write-up — we would just like to know who is reading. One form opens every case study on the site for the next 30 days.

If this is your problem too

Start with a fixed-price pilot.

Two weeks on a real piece of your backlog, priced before we begin. You keep the output whether or not you continue.

Related work

Review management · Conversational AI

Chatbots their customers build themselves

Healthcare · Synthetic data