Some of the biggest and most powerful companies in technology are racing each other to make machine learning easier to use, and DimensionalMechanics, a Bellevue-based startup, has $1.2 million in new funding and a new version of its NeoPulse framework ready for this market.

DimensionalMechanics co-founder and CEO Rajeev Dutt

Companies use DimensionalMechanics’ NeoPulse Framework to build their own AI-powered applications and products, relying on its “oracle” to automate the production of machine-learning models that will work best for their data and business models. Later on Monday, the company is releasing the 2.0 versions of two key components of the NeoPulse Framework, the NeoPulse Modeling Language and NeoPulse AI Studio.

“Our goals remain the same: giving developers the ability to make custom models,” said Rajeev Dutt, founder and CEO of DimensionalMechanics. “We are really aiming to be the single platform for all AI needs,” he said, explaining that the company targets both newbie AI developers as well as experts who want to take advantage of a lot of custom bells and whistles.

The NeoPulse Modeling Language 2.0 is now “Turing complete,” Dutt said, meaning that the language can now be used to develop anything that other major programming languages are capable of developing. The language also now supports new types of data such as audio and video, while the first edition was designed just for static image recognition, he said.

The new version of the AI Studio upgrades the key part of the software — the “oracle,” as the company calls it — that takes a customer’s data and decides what machine-learning models are most applicable to that data. The new version uses spectral optimizations to “improve the range and scale of the problems we can support,” Dutt said.

DimensionalMechanics also scored new funding recently to help build out the company and product, raising $1.2 million from undisclosed angel investors as part of a larger $5.5 million round that also included the conversion of some outstanding debt to preferred stock. The company raised $4.7 million in 2016 from another group of angel investors, and is calling the new funding a follow-on to that round, as opposed to a Series B round.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.