DarwinAI's Generative Synthesis technology – the byproduct of years of scholarship from the University of Waterloo – uses AI itself to observe a neural network as it is being trained and then generates a number of new and highly optimized versions of said network.

The result are new and entirely unique networks that are not only significantly smaller than the original, but infer faster without sacrificing functional accuracy. Moreover, the understanding garnered by the engine enables explainable deep learning, whereby the platform can illuminate how a network is reaching its conclusions across an entire dataset or a singular inference.

The benefits of DarwinAI's Generative Synthesis platform are threefold:

1. Accelerate deep learning development by collaborating with AI during the design process.

2. Reduce computational requirements by generating highly optimized models (particularly useful in deploying deep learning at the edge).

3. Understand how a neural network is reaching its decisions – important in both diagnostic and regulatory contexts.

In addition, the Darwin platform is complementary to low-level hardware optimization techniques found in toolkits such as Movidius, making it is essential for deploying deep learning applications to Intel chipsets.

Industry recognized scholarship

DarwinAI's proprietary IP is the product of years of scholarship from the University of Waterloo. Led by Professor Alexander Wong – a founding member of Waterloo's AI Institute and Canada's research chair in Artificial Intelligence – the team's work has been featured in TechCrunch, VentureBeat, Synced, The Next Platform, InfoWorld, JCVIS, IEEE Access and Forbes. Dr. Wong has authored over 450 peer-reviewed scientific papers, and is a recognized leader in the fields of artificial intelligence and computational imagining, a distinction for which he was named a member of the Royal Society of Canada.

DarwinAI is an alumni graduate of the 2018 Creative Destruction Lab (CDL) program, was recently included on insideBigData's list of top 50 impactful companies for Q1 2019, and was selected as one of seven finalists (out of 4500) for the Hello Tomorrow global summit in Paris in March.

The company's technology has been used in numerous industrial contexts including Autonomous Vehicles, Consumer Electronics, Aerospace and Military, Security, and Financial Services.

Please click the ‘Services’ tab below for some screenshots of the Darwin platform.

DarwinAI Video

The Generative Synthesis platform consists of a web-based interface that integrates with popular deep learning frameworks such as TensorFlow. The platform can be deployed on-premise or made available via a cloud-based setup.

The workflow is straightforward:

The user inputs a model and specifies performance requirements for the generative process.
DarwinAI Figure 01
The platform generates numerous models that adhere to these requirements, which are both smaller and faster than the original.
DarwinAI Figure 02
Our visualization tools pinpoint any problems and bottlenecks.
DarwinAI Figure 03

DarwinAI Figure 04
The new and optimized model can be deployed to Intel chipsets using existing tools like Nervana / Movidius.