Ayasdi: Disrupting Industries with Enterprise Artificial Intelligence

Ayasdi is a pioneer in enterprise-grade artificial intelligence. The company operates in three key markets; financial services, healthcare and the public sector. Ayasdi is focused on building intelligent applications to cover a range of mission-specific challenges such as anti-money laundering, fraud, cyber, clinical variation management, population health and customer intelligence. These applications present different user experiences but all draw from Ayasdi’s proprietary machine learning platform.

Ayasdi’s mission is to extract value from the world’s complex data. The company was incubated at Stanford’s mathematics department and is based on a technology called Topological Data Analysis, which is derived from a powerful branch of mathematics called Topology. Topology seeks to reveal the shape of data. By understanding the shape, one can discern a tremendous amount about the data.

This work was initially funded by Defense Advanced Research Projects Agency (DARPA) and later by venture capitalists. To date, the company has raised over $100M through three rounds.


The Influential Leaders of Ayasdi

Ayasdi has three Co-founders, all Ph.D.’s in mathematics at Stanford. They are Gurjeet Singh, CEO; Gunnar Carlsson, President; Harlan Sexton, VP of R&D. Gunnar Carlsson is considered to be one of the leading mathematicians alive today and his work in the field of topology has earned considerable accolades. It took Gurjeet Singh, one of Dr. Carlsson’s graduate students, to turn that math into software. Gurjeet’s combination of math and computer science enabled the powerful mathematical framework to be rapidly and easily applied to a number of different enterprise-grade problems.


Solving Real-World Business Problems

Advances in AI over the past few years have focused on the ability to build better predictive models using a variety of techniques, not the least of which is deep learning. These approaches, however, require massive data sets and restrict their utility for enterprise-grade problems. Complicating matters further is that the vast majority of enterprise data is unlabeled creating an additional class of challenges.

The future of prediction lies upstream in the analytical process – in the unsupervised learning techniques of segmentation, anomaly detection and hot spot detection. Here, unencumbered by labels, or even the requirement to know what you are looking for, lies in the keys to improving the performance of prediction. Ayasdi has pioneered the application of unsupervised learning techniques to enterprise problems, delivering enormous impact on some of today’s toughest enterprise use cases in finance, government and healthcare.These include intelligent segmentation for anti-money laundering, intelligent reserving for liquidity management, population health, clinical variation management, text corpus navigation and others.


Driving Innovation with Machine Learning

The pace of technological innovation is accelerating and is much evident today. What is profound is that so many different areas are drawing from the same base – a revolution in machine learning. The result is cross-pollination at an exceptional scale. ML drives innovation in robotics, which generates data from an IoT perspective, requiring cloud processing from an economic perspective and demands new computing paradigms like GPUs. It is an ecosystem and at the center is machine learning.

The result is that we live in a world where technology providers and technology savvy incumbents are surging ahead of their competitors. While risks, regulation and ethical questions are real – we expect continued acceleration in the pace of change.


Distinctive Services That Set Ayasdi Apart

Ayasdi is somewhat unique in that it is a company based on a mathematical framework – that of topology. This powerful approach was developed over decades at Stanford and seeks to discern the shape of data, which is an exceptionally powerful mechanism for understanding patterns, relationships ultimately driving the creation of predictive models.

As a result, Ayasdi has built almost all of its software in-house. The company has benefited from its relationship with Intel in optimizing compute. Further, Ayasdi partners with a number of firms on the change management side of the equation – an area that is grossly overlooked when operationalizing AI.


Tackling Business Challenges

The challenges faced by the company can be described in both technical terms and in operational terms. The underlying power of the platform is widely recognized, however, two key technical elements have proved challenging. First, optimizing performance and scaling the algorithms to deal with exceptionally large and complex data sets. Second, creating an application infrastructure to broaden the number of deployment configurations.

First, scaling math is very difficult, and comes with non-obvious trade-offs. Ayasdi has invested tirelessly to create a platform that is computationally efficient as well as performant. Second, the creation of an application infrastructure is key as the way AI goes to market in an enterprise is not through models, but through applications. Building the application layer is critical to the operational success of the technology.

Both of these lessons were hard won.


Recognition and Achievements

As a leader in the space, Ayasdi has earned broad recognition for its work. Some of the notable ones include:
1. Fast Company World’s Most Innovative Companies (2014, 2015)
2. Forbes FinTech 50 (2018)
3. World Economic Forum Technology Pioneer (2015)
4. Chartis Top AI Company in Financial Services (2016, 2017)
5. CB Insights Top 100 Companies in AI (2017)


Future Road map

Aysadi has exceptional ambition. The company sees itself as the future operating system for the intelligent enterprise. That means its platform is powering hundreds, if not thousands of applications inside an organization.

This aligns with Aysadi’s vision for the industry as a whole – that of an enterprise defined increasingly by intelligent applications.