Customer cases

Below you can see some of our customer cases.
They are all built on the Alvíss AI™ platform.

CONSTRUCTION

Quali-Drone

AI based surface damage detection

MEDTEC

Leo Innovation Lab

AI based skin cancer identification

INVESTMENT

AI Alpha Lab

AI based investment optimazion

INFRASTRUCTURE

Uno-X

AI based sales forecasting

CONSTRUCTION / VISION

QUALI-DRONE

AI based damage detection

Challenge
Bladt Industries is the market leading manufacturer of large offshore substations for the oil, gas and windmill industry. As the constructions are exposed to a harsh environment out at sea, it is crucial that the surface seal is 100% correct upon delivery. Until now, quality inspections have been performed manually, which is dangerous, time consuming and expensive.

Solution
In collaboration with specialists in drone technology, precision measurement and digital twins, Desupervised has developed AI models for a solution that automatically inspects the constructions, identifies surface damage via advanced AI based analysis and visualizes the results in a digital twin.

Business value
The new inspection solution will significantly reduce the time consumption and costs of surface quality control of the structures, increase the inspection quality and capacity, as well as provide documentation of identified damage and the subsequent repair of these.

MEDTECH / DIAGNOSTICS

LEO INNOVATION LAB

AI based skin cancer detection

Challenge
Leo Innovation Lab wanted to develop an App giving users the opportunity to diagnose skin diseases based on a photo of skin. Being part of a healthcare company, it was imperative to Leo Innovation Lab that the suggested diagnostic results were within a specific certainty range, given by domain knowledge.

Solution
Using Alvíss AI built on Bayesian Deep Probabilistic Programming, accurate and uncertainty aware models were developed and tested for identification of skin diseases.

Business value
Alvíss AI enables and quantifies uncertainty in the diagnoses which is imperative for out of sample predictions (for example photos not containing skin)

Because Alvíss AI requires much less data and extracts more information per data observation, this was a major contributor in estimating the uncertainty. This is a major improvement in the quality of the actual suggested diagnostic results compared to results based only on maximum likelihood where over confident predictions are abundant.

ASSET MANAGEMENT / RISK MANAGEMENT

AI ALPHA LAB

AI based investment

Challenge
AI Alpha Lab has created an investment AI that operates on all markets in the world. However, predicting the return of every individual asset in a market using machine learning is extremely challenging since models fit well but generalize poorly

Solution
While predicting the absolute return accurately is indeed nearly impossible, predicting the risk associated investing in each asset can be done. This requires taking all market data into account at the same time.

Alvíss AI can use every single price movement, macro economic factor as well as fundamental analysis to create posterior densities over the return for each asset coupled with its own specific uncertainty.

Business value
Optimizing the ROI/risk on investing in assets

INFRASTRUCTURE / SALES

UNO-X

AI based sales forecasting

Challenge
Predicting future sales for a gas station on an hourly basis is a complicated task. It involves many dependent factors such as pricing, weather, season, location and road works etc. Without accurate modelling stations risk of running out of fuel before the next supply is delivered – thus losing revenue and customers to competitors in the short and long term.

Solution
Based on the Alvíss AI platform, Uno-X has implemented AI models in its sales forecasts for each station. Each prediction comes with a quantification of uncertainty, allowing the logistics department to adjust expectations accordingly.

Business value
Minimizing the risk of lost revenue, customer churn and negative reputation.