We developed a solution based on Computer Vision and AI for the real-time extraction of high level contextual information from video stream data, for a leading player in the gaming industry.
We developed CI/CD -ready data processing pipelines for video data for the purpose of Quality Assurance of hardware/software products, for a leader in the gaming hardware industry.
This is foldAI’s internal project. We are developing an AI-powered ecosystem monitoring solution, spanning from the design and production of our own sensing Edge devices, to the back- and front-end development of the software and data infrastructure. Discover more at https://fold.ai .
To help a customer facilitate rapid prototyping with Machine Learning algorithms we delivered a full stack machine-learning-pipeline optimized for multi stream time-series data. This enabled the customer to create and deploy Machine Learning algorithms within hours, while increasing predictability and reproducibility in their development workflow.
We also support clients with extending and refining their existing infrastructure to integrate Machine Learning workflows, as well as provide advice on data strategy, and generalist data science support. As an example in this area, we are currently working with Kollektiv, a startup providing a platform for personalized training for triathletes.
Detection of biometrics and tracking of motion activities requires a combination of expertise from signal processing and machine learning applications. We have experience in creating and deploying several such data driven algorithms for customers producing wearable devices.
Our client sought to improve the total quality of their production line. To achieve this they needed readily digestible insights into their KPIs. Working with the client we developed algorithms for deployment within their production and testing infrastructure. This enabled us to deliver an interactive dashboard where the client could easily view metrics for a given period, such as the nth -pass yield across each workstation, and most significant error codes.