Crushing $100M/yr Overhead

SITUATION

A Seattle-based tech giant manually processed identification documents for over 500k new employees globally. This created a staggering $100M YOY overhead cost and a strong desire to optimize. We were looking at a technology solution that relieved the problem by only about 10%. We were scrambling for an answer to the problem and to maintain the interest of this strategic partner.

HYPOTHESIS

I noticed a few gaping holes in the business process from which this $100M overhead was leaking. If we could deliver on the following bullets, we should be able to reduce this overhead by at least 75%:

• Place the onus of QA on the applicant

• Set applicants up for success

• Dig deeper into the client’s acquired technology

SOLUTION

Left with only 10min after an underwhelming 50min demo, I presented the hypothesis, including proof of concept materials, and had our tech giant literally speechless.

Place the onus of QA on the applicant
Optimized the business process by including an additional step for the incentivized applicant to double-check their information was correct. Applicants were given three opportunities to capture their documents and ensure their information was correct before it landed in the queue of an HR Generalist.

Set applicants, and technology, up for success
We surfaced an acquired set of features that would straighten, desaturate, and increase the contrast of applicant images to ensure their ability to be OCR’d with the greatest possible efficacy.

Dig deeper into the client’s acquired technology
In addition to image optimization, I found a set of features that functioned nearly identically to AWS Rekognition. This software is amazing in its own right and ensures that applicants having to resubmit photographs would be kept to a bare minimum very strict bare minimum.

RESULTS

Early use in Canada is proving to increase the efficacy of the solution from 10% to 95%, nearly 10x crushing the aforementioned $100M/yr overhead.