Sagra Technology needed to audit pharmaceutical products on pharmacy shelves fast and accurately, across 25 near-identical packs that differed only by the number of doses. Working with Neurolabs, it went from onboarding to project delivery in seven days, a fraction of the market standard of a few weeks, at 98.3% detection accuracy.

Founded in 1998, Sagra builds mobile software for sales, marketing and analytics, and credits much of its success to speed: it works differently from the trend and stays a technological step ahead of the market. As an innovator, it needed a partner that could work in the demanding pharmaceutical space, cut its clients’ onboarding time and deliver image recognition results in the moment.

The challenge: telling near-identical products apart, fast

Sagra had 25 practically identical products to audit, distinguished only by the dose count on each pack. It needed to give sales reps results during the pharmacy visit itself, while holding the highest level of detection accuracy. Traditional image recognition struggled on both counts: minor differences in packaging are hard to distinguish and preparing a model the classic way is slow and data-hungry.
‍

We have been embedding traditional IR technology in Emigo, our SFA solution for some time, but quickly realised we need to provide results faster and run the IR service for customer in days not months. The time-consuming process of preparing the AI recognition model was not a viable solution for us.

We needed a solution that could quickly onboard customers and create models significantly faster but with maintaining in the same time the highest possible accuracy, and of course being able to integrate seamlessly with our Emigo SFA System. Neurolabs’ ZIA technology has been a true game-changer for us. Its synthetic data approach to IR as well as the way of model creation and new SKUs onboarding has revolutionised our operations!
‍
Sagra Technology

‍

The difference: training on 3D digital twins

Classic image recognition needs a large amount of real data, collected and then annotated by hand, box by box, for every product. It is slow to gather and slow to train. Neurolabs works differently. ZIA (Zero Image Annotation) generates synthetic images and 3D digital twins instead of annotating real photos, which is faster and has proven more robust and reliable than real data. For Sagra it delivered 98.3% product detection accuracy. That is typical: ZIA averages 95% visual detection accuracy from the outset, rising above 98% for specific categories. To remove any doubt, Neurolabs also offers an independent third-party quality audit to prove the accuracy meets what was agreed.

Because it trains on twins rather than photos, Neurolabs ZIA learns a product from millions of positions and angles. So it holds accuracy in the real world, where packs get picked up, rotated, opened or moved into poor light and where near-identical products would defeat traditional IR. For Sagra’s 25 SKUs, ZIA quickly learned to tell them apart by dose alone. It also let Sagra shift from reactive to proactive: new products and packaging changes could be uploaded before they even hit the shelf.

Synthetic recognition in action

To start, Sagra sent PDF images of the packaging labels for all 25 SKUs. Neurolabs used ZIA to build synthetic 3D assets of each, then trained and tested them in a virtual environment. Training took about five days, enough to teach the algorithm the subtle differences between labels, after which auditing scenes could be built from synthetic data. Sagra’s own tests found the algorithm detected products at 98.3% accuracy, with results in about five seconds on average, well beyond its expectations. Integration was quick too: Sagra connected ZIA to its existing stack via cloud APIs. ZIA is built to onboard fast, adding a CPG customer in as little as a day and reaching a time to market of one week for up to 1,000 SKUs.

‍

Neurolabs not only managed to deliver a viable solution to our challenge in an unprecedentedly fast and efficient manner, but they also did so with near-perfect accuracy. We faced some challenges with some pictures (blurs, pictures from above, glares, products covered with shelving elements) but we are well aware of them and together we are working to address them (e.g. providing alternative predictions, improvement of the data collection process).
‍
Sagra Technology

‍

The partnership ahead

After a successful first stage, Neurolabs and Sagra are continuing to work together to speed up the adoption of image recognition in Poland. With a focus on both pharma and fast-moving consumer goods, the shared goal is to serve customers of all sizes across both sectors.

Auditing near-identical products, or need results in the store? Talk to our team today.