UpStride
New computational paradigm to improve accuracy, data efficiency and power consumption of neural networks.
| Date | Investors | Amount | Round |
|---|---|---|---|
| - | investor investor investor investor investor investor | €0.0 | round |
N/A | Acquisition | ||
| Total Funding | 000k | ||
Related Content
UpStride was a specialized deep-tech firm established in Paris, France, in 2018 by co-founders Gary Roth (CEO) and Wilder Lopes (CTO). The company focused on enhancing the efficiency of artificial intelligence models, particularly in the domain of computer vision. UpStride developed solutions to reduce the cost and data requirements for training AI models without necessitating changes to the underlying code, neural-network architecture, or datasets.
The core of UpStride's offering was a proprietary technology centered around 'Hycor,' a novel data type specifically engineered for deep learning that leverages geometric features within data to improve information representation. This technology was delivered through APIs designed for easy integration into existing AI development pipelines, such as those using TensorFlow, Nvidia's cuDNN, and Intel's OneAPI. The primary benefit for its clients—AI developers and data scientists across sectors like retail, automotive, and healthcare—was the ability to improve model accuracy and reduce power consumption. In 2020, the company released an open-source image-classification API to simplify common computer vision tasks for deep learning experts.
The firm's trajectory culminated in its acquisition by Contentsquare, a prominent digital experience analytics platform, on July 13, 2021. This strategic acquisition brought UpStride's team of 14 deep-learning engineers, including several PhDs, into Contentsquare's R&D division to bolster its AI capabilities. CEO Gary Roth transitioned to a strategic role on Contentsquare's operations team following the acquisition. The deal was intended to accelerate Contentsquare's AI innovation roadmap, particularly in areas like intent-based optimization and predictive analytics.
Keywords: UpStride, deep learning, machine learning performance, AI model training, computer vision, Hycor, API, data efficiency, neural networks, Gary Roth, Wilder Lopes, Contentsquare acquisition, AI efficiency, image classification, data training solutions, AI optimization, deep-tech, artificial intelligence, model accuracy, reduced data