UPDATED 09:00 EDT / OCTOBER 23 2018

AI

Sisense launches Hunch, an AI engine for bringing complex analytics to the edge

Sisense Inc. believes there’s a better way to process information at the edge.

The DFJ-backed analytics startup today unveiled Hunch, an artificial intelligence engine that it claims can condense large datasets into a size manageable even for connected devices. The software does so by creating a neural network that acts as a stand-in, taking up only megabytes of space for every terabyte of information being substituted.

Compressing large analytics workloads to fit a hardware-constrained environment is nothing new. One common method is to create a summary of the records being analyzed, while another approach is to extract a sample and use it to generate a less resource-intensive statistical approximation of the dataset. Yet while they offer certain advantages, both methods require tradeoffs in areas such as accuracy that Sisense said are mitigated to a large degree by Hunch.

According to the startup, the engine works by extracting patterns from a dataset and generating a large range of queries to simulate those that business users might run against the information. From there, Hunch instills this knowledge into a lightweight neural network that can answer queries offhand without having to comb through the original records.

Removing the need to run a full-blown query every time a user requires a certain piece of information saves a lot of time. According to Sisense, AI models generated with Hunch can provide submillisecond response times with accuracy of upwards of 99 percent. Plus, the lightweight nature of these neural networks makes it possible perform analyses on relatively small devices.

“Once a Sisense Hunch neural network learns data, it doesn’t need any ongoing access to the complete underlying data set, allowing it to achieve lightning-fast, analytical query responses, with minimal costs, while maintaining complete data privacy,” Sisense Chief Executive Officer Amir Orad said in a statement.

Hunch does have to make some tradeoffs to provide these advantages: The neural networks don’t provide the same level of detail as the datasets on which they’re based. In situations where users require a higher level of granularity than what a neural network can handle, the engine routes requests to the system containing the original information.

Hunch is already seeing some early traction in the market. Sisense said that the engine been adopted by a publicly traded electronics maker to optimize quality control operations at a plant, while several other companies are currently testing it.

The introduction of Hunch marks Sisense’s first product launch since its $80 million funding round last month. The company has raised about $200 million in total and claims more than 1,000 customers, including General Electric Co., VMware Inc. and the Nasdaq Stock Market.

Photo: Pixabay

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU