Klepsydra Technologies provides patented edge AI software solutions optimized for high-performance, low-latency inference in demanding sectors like aerospace and robotics. Their acceleration technology achieves significant gains in throughput and power efficiency through two-dimensional parallelization and lock-free programming. The unified API ensures hardware independence, allowing clients to deploy optimized AI across diverse edge computing platforms.
Funding
$2.6M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
SPFounders
Product
Problem
Edge computing faces challenges in processing large data volumes on resource-constrained devices, leading to data loss and delays in critical applications such as autonomous vehicles, robotics, and aerospace. The increasing demand for real-time data processing at the edge outpaces the computing power of these devices.
Solution
Klepsydra Technologies provides a Software Development Toolkit (SDK) and software libraries designed for high-performance data processing and AI at the edge. The platform optimizes algorithms for small edge devices, enhancing performance by processing up to ten times more data with significantly reduced power consumption. Inspired by high-frequency trading, Klepsydra leverages advanced multi-threading techniques to enable the development of high-performance applications. The Klepsydra AI inference engine allows pre-trained AI/ML models to run on edge devices in real-time, processing more data while reducing power consumption. The Robotic Toolkit optimizes robotics applications, ensuring sensor data is processed without losses and reducing CPU power consumption.
Target Audience
Klepsydra serves industries such as aerospace, robotics, automotive, and IoT, focusing on clients who require high-performance, real-time data processing and AI capabilities on edge devices.
Features
- Klepsydra SDK optimizes algorithms for small edge devices, reducing processor power consumption.
- Klepsydra AI enables running pre-trained AI/ML models on edge devices for real-time applications.
- Klepsydra AI processes up to six times more data than competitors, reducing processor power consumption by up to 50%.
- Klepsydra GPU Streaming provides efficient GPU programming capabilities using a C++/CUDA interface.
- Robotic Toolkit includes ROS, ROS2, and OpenCV plugins to optimize robotics applications.
- Acceleration technology combines two-dimensional parallelization with lock-free programming.
- Autotuning tool optimizes AI inference for various constraints.
- Unified API eliminates processor lock-in.