DimensionalMechanics
Democratizing AI for Building AI tools

Despite all the spiraling advancement in the tech world, it is still a long shot for many enterprises to get up and running with the newest advancements. Machine learning is no exception. Difficulties in finding skilled employees, organizing data sets, timely and customized development efforts, and steep price points result in a massive gap in the accessibility of such breakthrough technologies. On the other hand, many machine learning solution providers work only around specific verticals, which takes away the opportunity for enterprises to test and identify re-usable approaches or applications that are best suited for them. Based out of Washington, DimensionalMechanics addresses these predicaments by democratizing their powerful machine learning stack of offerings that can be molded to work about datasets that enterprises already own. The company makes AI more accessible to organizations of all sizes, by dramatically reducing conventional barriers to entry.

The company’s NeoPulse AI Studio is a powerful AI tool that is capable of automating AI model creation. DimensionalMechanics also developed an intuitive domain specific language called NeoPulse Modeling Language (NML) to simplify and speed AI development. “There are no languages out there that were particularly designed for deep learning. With programming languages like Python, there is still a need to write a ton of code; whereas, in the case of NML, that effort can be reduced by more than 85 percent,” explains Rajeev Dutt, President and CEO of DimensionalMechanics. Another crucial component of the DimensionalMechanics solution stack is the portable inference model (PIM) which is a containerized neural network AI model that gives developers the powerful ability to move their AI studio projects across machines to on-premise hosts and the cloud. NeoPulse AI Studio was recently launched on the AWS marketplace with one-click installation feature. The ability to be queried during runtime adds to PIM’s flexibility. Additionally, with the NeoPulse Query Runtime, organizations can allow any applications in the enterprise to access AI models using REST APIs.

DimensionalMechanic’s novel “AI to build AI” solution means developers need not be deep learning experts themselves in order to benefit.

Our roadmap will always be focused on making machine learning an integral part of every enterprise’s decision-making process

Their developer portal has ample resources for even regular developers to train themselves, design, and optimize AI models. But, for companies that do have deep learning teams, DimensionalMechanics enables them to work much faster, and finally use a set of common tools to approach vastly different internal business challenges. Apart from its robust and flexible stack and speed of deployment, one of the biggest differentiators of DimensionalMechanics is pricing, which Dutt notes, is a fraction of other players in the space.

The media and entertainment industry, DimensionalMechanics initial target market, have used some pre-built models created in NeoPulse AI Studio in a variety of interesting ways. One client placed their model for ranking digital images in the top 3 solutions in a head-to-head comparison with about forty similar offerings in the space— but it was done at a fraction of the cost and time of all competitors. A news headline optimization model was able to increase click rates by up to 200 percent in a test by a US news organization. In another instance, a large network infrastructure provider was able to use NeoPulse AI Studio to predict network traffic 90 percent of the time, and create an AI model that could monitor the network and simultaneously train itself. The company’s horizontal framework can be used across any industry. Beyond media and entertainment, DimensionalMechanics has recently generated interest from companies in the infrastructure, medical, and education space. “We see new applications being developed in some very creative ways using NeoPulse. And this is exactly what we intended to unleash and empower. Our roadmap will always be focused on making machine learning an integral part of every enterprise’s decision-making process,” concludes Dutt.