Abacus.AI Secures $ 50 Million Funding, Examines Computer Vision and Hybrid AI Models


Starting machine learning operations Abacus.AI Inc. is in the news today after raising $ 50 million in a new round of funding and launching platform updates that include support for hybrid deep learning models and a new offering of computer vision as a service.

The Series C funding round was co-led by Tiger Global, Coatue, Index Ventures and Alkeon Ventures and brings the total amount of Abacus.AI raised to $ 90.3 million.

Abacus.AI has been able to attract those millions through its enterprise-wide MLOps platform that businesses of any size can use to deliver real-time data from online purchases, interactions on social networks, “Internet of Things” sensors and more. Then, they can process that data, transform it, and use it to create deep learning-based artificial intelligence models capable of generating contextual predictions in real time.

Until the arrival of Abacus.AI, only a few large companies with hundreds of thousands of dollars to invest in such companies were able to build such complex models.

Thus, the Abacus platform effectively brings deep learning to the masses. It enables an MLOps discipline that aims to implement workflows by promoting more collaboration and communication between data scientists and developers.

It provides all the components needed to build complex AI systems, with easy configuration of data pipelines, data cleansing and transformation tools, model training and hosting, monitoring and monitoring capabilities. explainability, etc. Users can train models using Abacus.AI’s own neural network architecture or use an open source framework such as PyTorch or TensorFlow.

The company has effectively created an “end-to-end AI service” that has the potential to become a one-stop-shop for all business AI needs, Andy Thurai, analyst at Constellation Research Inc, told SiliconANGLE.

It’s a powerful platform that just keeps getting better. With today’s release, Abacus.AI adds support for “hybrid models” that can generate predictions from a combination of language, vision, and tabular data. According to the company, this is a new technique that relies on neural network architecture research combined with pre-trained integration for image and language data to extract intelligence.

This is a promising development as existing AI and machine learning tools typically only work with structured tabular data, which means unstructured data such as photos, videos, and call transcripts cannot be used to train models. With Abacus.AI’s hybrid models, this unstructured data will no longer be wasted. Instead, it can be used to build models that are more complex and perform better than ever before, combining tabular, textual, and vision-based data.

Abacus AI co-founder and CEO Bindu Reddy (pictured) told SiliconANGLE that hybrid AI models are a logical next step, as it’s actually quite easy to use unstructured data to train the ‘AI using modern deep learning techniques. The problem is, most organizations traditionally use classic tree-based detection models that can only work with structured data.

Abacus AI has created new deep learning techniques that can extract signals from language and image-based data with relative ease. As a result, companies will now be able to create models that significantly outperform conventional tree models, she said.

“Data science teams can now create complex models and extract powerful information in hours using whatever data they have, whether tabular, textual or visual,” Reddy explained. “For example, you can predict the closing price of homes based on unstructured data such as listing description and photos of the home, as well as structured tabular data including number of bedrooms, bathrooms, and bathrooms. baths, etc. This is done by combining all of this data and using the Abacus.AI predictive workflow to generate a hybrid predictive model that combines all types of data. It is a very powerful way to extract information from all the data available.

Additionally, Abacus.AI enables businesses of all sizes to easily implement computer vision use cases such as image detection, classification, and segmentation. The new Computer Vision-as-a-Service offering adds to Abacus.AI’s existing support for predictive analytics, personalization, anomalies, forecasting and language. Now, users will also be able to develop bespoke deep learning-based computer vision models, Reddy said.

Going forward, Reddy said Abacus.AI plans to add support for additional common business use cases in Vision AI and continues to invest in new neural network techniques for verticals. such as customization, forecasting, anomaly detection and linguistic hybrid models.

“We are also expanding our research areas to include reinforcement and meta-learning,” she said. “Finally, we continue to deepen our ML and DLOps modules which include a real-time feature store, model drift, observability, and a vector matching engine.”

Photo: Abacus.AI

Show your support for our mission by joining our community of Cube Club and Cube Event experts. Join the community which includes Amazon Web Services and Amazon.com CEO Andy Jassy, ​​Dell Technologies Founder and CEO Michael Dell, Intel CEO Pat Gelsinger and many other luminaries and experts.

Leave A Reply

Your email address will not be published.