On the Radar: Cortex Labs, RadicalBit, Qeexo
Welcome to “On the Radar,” a new Datanami feature that will showcase new companies and technologies that we think data practitioners may like to know about. In this week’s debut feature, we focus on three companies that have just “hit our radar,” so to speak: Cortex Labs, RadicalBit, and Qeexo.
Cortex Labs helps organizations put machine learning models into production on the cloud, much like AWS SageMaker. Its software as a service (SaaS) offering raises the abstraction level on machine learning deployment, thereby reducing the need for data scientists and machine learning engineers to know the ins and outs of operating core components of the cloud stack, such as using containers, Kubernetes, and Nvidia GPU drivers.
Cortex currently supports AWS, and gives users the choice of GPU and CPU environments. When a user is ready to deploy his ML model, he hands it over to the Cortex environment, and Cortex is then responsible for packaging the model as a container and ensuring that it’s scalable enough to meet demands on AWS. Users are given some dials to turn regarding performance and efficiency, the company says.
The Oakland, California company develops its software in the open, and makes its source code available on its GitHub page. Company officials say there currently is no standard software or approach for deployment of ML models into product, and they hope that Cortex will fill that role, which is why they’re making it available as open source. Corext works with an array of ML frameworks, including TensorFlow, PyTorch, nscikit-learn, XGBoost, and others. Stay tuned for an upcoming story on Cortex Labs on Datanami in the near future.
RadicalBit develops software designed to help customers move into the streaming data age. The window of opportunity for acting upon data is shrinking, and RadicalBit hopes to help customers take advantage of data soon after its generated with its end-to-end DataOps platform.
RadicalBit starts with core technologies like Apache Kafka, Apache Flink, and Apache Spark, and then extends these components by providing user interfaces that are designed to make it easier for customer to work with these technologies for developing and maintaining data pipelines and real-time machine learning applications.
The Milan, Italy-based company has served customers in the logistics and automotive industries, including development of a IoT application that tracks vehicle movement and serves the data in real time. The company has customers in Europe as well as the United States. Watch these pages for more from RadicalBit soon.
Qeexo is a machine learning startup that develops AutoML software specifically tuned to automatically detect patterns in sensor data. Models developed by Qeexo AutoML (which it touts as a “one-click, fully automated platform”) are designed to run on low-power edge devices that are sensitive to network latency, such as mobile apps, IoT devices, wearables, and automobiles.
Arguably its biggest win to date has been Huawei, which used Qeexo’s FingerSense technology, which can differentiate among the different ways that users touch their mobile devices. Hauwei rebranded the technology as “Knuckle Sense,” and the tech has been deployed to 200 million devices.
The Mountain View, California was spun out of Carnegie Mellon University’s Human-Computer Interaction Institute in 2013, and has raised $7.4 million in venture capital. Qeexo launched its AutoML product late last year. Stay tuned for more about this company at Datanami.
Got a tip for On the Radar? Drop us a line at [email protected]