The startup develops deep learning technology that integrates machine learning capabilities into electronic devices and robots, enhancing their computational power and connectivity. This technology enables the creation of smart products that perform tasks efficiently, improving user safety and convenience.
Funding
$8.9M 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.
Founders
Product
Problem
Many AI solutions for embedded devices require significant computational resources and cloud connectivity, leading to increased costs, latency, and privacy concerns. Existing solutions often struggle to balance accuracy with efficiency, limiting their deployment on low-power devices.
Solution
Plumerai provides a complete software solution that enables highly accurate AI on resource-constrained embedded devices, eliminating the need for cloud processing. Their technology combines optimized inference engines with tiny AI models, enabling advanced features like familiar face identification, people detection, and object recognition to run efficiently on microcontrollers and low-power SOCs. By processing data locally on the device, Plumerai's solution reduces latency, enhances privacy, and lowers costs associated with cloud computing and bandwidth usage. The software is designed for easy integration with existing camera systems and other IoT devices, offering a balance of accuracy and efficiency.
Target Audience
Plumerai's primary customers are developers and manufacturers of smart home cameras, video doorbells, security cameras, video conferencing systems, and other IoT devices seeking to add advanced AI capabilities to their products while maintaining low power consumption and respecting user privacy.
Features
- Optimized inference engine for Arm Cortex-M, Arm Cortex-A, RISC-V, and x86 processors
- Tiny AI models trained on a dataset of over 30 million images and videos
- Support for INT8 quantization for reduced memory footprint and faster inference
- Complete software solution for smart home cameras, including familiar face identification, stranger identification, people detection, vehicle detection, and animal detection
- Advanced motion detection that filters out false positives caused by rain, snow, or lighting changes
- Multicam re-identification for seamless tracking of individuals across multiple video streams
- C++, C, and Python APIs for easy integration into existing systems
- Compliance with privacy laws, including GDPR, CCPA, and BIPA