Enterprise AI flywheel effect drives scalable use cases

The “flywheel effect” is here to stay in the rapidly evolving world of enterprise AI, driving collaboration between key industry players to build scale and resilience.

The “flywheel” business concept was first popularized by author Jim Collins in his book Good to Great, which explains how small, consistent efforts to build momentum over time can lead to breakthroughs and sustained growth. In the world of AI, measures of this progress can be seen more fully through initiatives such as collaborations involving Hewlett Packard Enterprise Co., SHI International Corp., and their partners.

TheCUBE, SiliconANGLE Media's livestreaming studio, spoke with leaders from HPE, SHI, and their partners at the recent SHI Fall Summit to highlight how strategies designed with resiliency, reliability, and long-term impact are critical to deploying AI at scale and producing meaningful results. Use cases for AI applications are beginning to emerge as enterprises incorporate AI into their infrastructure to integrate data, reduce risk, and build governance into system design. According to Tyler Webb, director of AI sales at SHI, the “flywheel effect” produces results that become more apparent over time.

“A lot of our customers are asking, 'What are the industry use cases that we're successful with, and where are other companies putting their time and effort?'” Webb told theCUBE. “We're talking about full-stack development, full-stack design, containers, everything below, everything above, and delivering solutions. Many of the surface outcomes will be tangible ROI from day one, but many of the benefits of these AI initiatives will be further down the line.”

This feature is part of SiliconANGLE Media's continued exploration of advances in AI innovation and cybersecurity resiliency. (*Disclosure below.)

Enterprise AI framework for the full stack

SHI employs a structured framework to help customers identify business challenges and envision how AI can create new opportunities. The goal is to prototype AI solutions using full-stack development and testing to see impact and measurable ROI.

“It's a framework called 'imagine, experiment, adopt,'” Webb explained. “We really start at the imagination stage where we understand the business problems our customers are facing and how AI can help them achieve the improvements and efficiencies they’re trying to achieve.”

This leads to use cases that actually impact specific processes and workflows. To accelerate this process, SHI launched the Digital AI Ambassador Development Platform in June in collaboration with Nvidia Corp. and HPE. The platform is designed to provide enterprises with expert-driven capabilities to quickly build and deploy custom AI-powered solutions. It is based on a combination of SHI's AI and Cyber ​​Labs services, pre-validated Nvidia AI blueprints, and HPE's private cloud AI.

“Digital AI Ambassadors are truly transforming the customer experience,” Webb said. “Whether you go to a retail store or go to the bank, in any situation, you're going to be interacting with a customer service agent. It creates a human-like experience, but it's done through an AI-powered avatar.”

A trusted repository for medical data

As AI use cases begin to expand, the technology's applications are also expanding into critical areas such as patient health. Sickbay, software used in hospitals to collect and provide physiological data at scale, is one example of how AI is steadily penetrating important centers of public life.

Sickbay, from Medical Informatics Corp., was developed in partnership with HPE to better manage data generated by patient-connected devices. This may include important information such as electrocardiograms and arterial blood pressure measurements, which can be difficult to find when needed in today's busy hospitals.

HPE infrastructure feeds AI models with massive amounts of medical data, creating a reliable repository that can deliver more accurate results.

“Understanding patient physiology at this sub-second granularity provides increasingly greater insights,” said Raajen Patel, Sickbay Executive Vice President of Innovation in Medical Informatics, in a discussion with theCUBE. “By collecting, accessing, and seamlessly distributing this data at scale and scale, [researchers] To obtain the desired subset of data, perform the analysis of interest, and find interesting information about the patient population. ”

Such insights can lead to improvements in how healthcare facilities deliver care. By introducing new standards of integrated data and monitoring to hospitals, healthcare teams can leverage AI to improve processes across the patient spectrum.

“This is a dataset, or a data stream that has not been available at scale before,” Patel said. “We're seeing all these patients with certain physiological symptoms, or we're seeing this particular type of patient in this geographic region. These patients have certain symptoms. We ask you to consider whether our policies and practices reflect the best care that we can provide.”

Deployment of smart city solutions

As the medical field advances, there are signs that AI will also impact the management of city services. City governments often face the challenge of modernizing their infrastructure and delivering more digital services while dealing with disconnected systems, tight budgets, and compliance demands.

HPE's Agentic Smart City solution is the result of the company's Unleash AI Partner Program, with support from SHI, Nvidia, and ProHawk Technology Group. By combining ProHawk's advanced AI-enabled computer vision with Nvidia's Metropolis stack, users can deploy low-latency solutions at the edge or on HPE's private cloud AI.

“The use cases go on and on,” Brent Willis, co-founder and chief operating officer of ProHawk, said in an interview with theCUBE. “That's the challenge and the benefit. It's part of the HPE Unleashed ecosystem, backed by HPE infrastructure and Nvidia infrastructure, so everything can work in a turnkey solution.”

ProHawk's computer vision management system includes an AI self-training element that emphasizes the growing flywheel effect. The company's technology restores pixels, essentially retraining the AI ​​model to become smarter at image recognition. This will allow cities to use ProHawk without requiring major reinvestment.

“The Achilles heel of all AI is data quality, and that’s the first thing we fix,” Willis says. “We restore the data that we receive so that all the other AI models and AI components work better and operate more efficiently. But what we do now is we restore the video that we receive because we have what we call an AI flywheel. It goes back again every 3 milliseconds and teaches the model, teaches the sensor, learns the model so that it can do all the identification.”

AppDev teams think holistically

Companies like ProHawk and Medical Informatics are working with HPE and SHI to create tighter connections between infrastructure decisions and how applications are designed and delivered. This is reshaping the application development or AppDev approach to delivering enterprise outcomes in today's IT world.

“HPE's focus on hybrid cloud, GreenLake, and platform-driven services, combined with SHI's integration and sourcing strengths, is forcing AppDev teams to think more holistically about portability (which our research shows is important to 20% of respondents), scalability, and lifecycle management from day one,” said Paul Nashawaty, Principal Analyst at theCUBE Research. “This may be a positive change, but it will require closer collaboration between AppDev, platform, and infrastructure teams.”

Nashawaty also noted that there is an opportunity to accelerate platform engineering practices on the development and operations and delivery side. If AppDev is brought into the discussion early and treated as a design partner rather than just a consumer of the platform, HPE and SHI can help improve speed, reliability, and alignment with business goals.

“HPE/SHI, when used well, can provide more self-service capabilities, a consistent environment, and better integration with CI/CD pipelines,” said Nashawaty. “That said, AppDev teams will need time and enablement to adapt, especially when moving from a bespoke setup to a more standardized platform.”

There is a central goal behind the partnership established by HPE and SHI. It's about creating easy ways for customers to move from AI experiments to tangible outcomes. The use cases are just the first wave of what is likely to become an established process for customers to quickly and securely build and deploy AI applications on reliable infrastructure.

“The goal is to make generative AI deployments repeatable and predictable,” said Fidelma Russo, chief technology officer and general manager of hybrid cloud at HPE. “With curated use cases and tested frameworks, we are taking the guesswork out of AI.”

(*Disclosure: TheCUBE is a paid media partner of SHI Fall Summit. Neither HPE, the sponsor of theCUBE's event coverage, nor any other sponsor has editorial control over theCUBE or SiliconANGLE content.)

Image: SiliconANGLE/Microsoft Designer