H2 Wellness uses mimik’s edge computing technology to slash its Amazon cloud services bill

Martin Banks Profile picture for user mbanks September 10, 2020
H2 Wellness has cut its monthly Amazon bill from $28,000 to $2,800 using mimik’s edge compute solutions.


The arrival of a company like mimik Technology, with its new approach to delivering edge computing capabilities, always raises a question – is there real world proof to its claims? An opportunity to talk with Hooman Fakki, Founder and CEO of US based healthcare services business H2 Wellness, gave the first clue that there is indeed. 

What makes it more interesting is that the company is not what one might expect as a typical edge computing-oriented operation. But many of its own customers can be, which makes it a good example to other businesses that might think that using edge computing models does not apply to them.

Edge computing, of course, is seen as the very front (often bleeding) edge of a business IT infrastructure, yet Fakki positions H2 Wellness firmly as a back-end engine providing digital health information and data management services to healthcare providers and advisors around the world:

They layer on their brands, and their own look and feel and their business flow, but we have a platform that spans across any type of health, wellness or fitness provider, globally.

The company provides those customers with the tools and information management to provide their end user customers help, guidance, support and services on their individual health journeys. This means the spread of customers can be quite wide, from doctors’ practices and hospital departments through to gyms and health advisory services. 

So the range of support H2 provides is equally wide, from collecting and collating individual patient data allowing doctors to track patient health status and make appropriate interventions, through to supporting advisory services with information resources and data that help them sell appropriate products to other service providers or end user consumers:

They need a digital platform to do it at scale and be effective. Their options are build it, which is very expensive and with a high chance of failure, lots of zigzagging to get it right. They can find something that is modular and licence it. And the third option is to try to find something off the shelf or open source, and those are riddled with a lot of problems. Nothing will fit.

Using the cloud, but less of it

By using something platform-based and SaaS-delivered, companies can build a fully-branded custom solution to fit their requirements at a competitive cost and what Fakki reckons is normally around a tenth of the time the other options can require. It also comes fully maintained, upgraded and extended as part of the subscription. This also gives them the ability to innovate on top of the platform to provide the specific services they need, either doing the work themselves or in collaboration with H2:

That innovation could be part of our existing roadmap, which we would accelerate, or that could be a joint project. Every option you can think of under the sun is available, our contracts allow that sort of flexibility.

Before moving to mimik, Fakki and his team conducted some extensive research on edge computing offerings, particularly as it is still early days in the technology’s development and, as he observed, it is quite easy to waste time on the wrong decisions. This also chimed with the company being at a key inflection point at which to make and implement such a move. He is aware that others have already acted and are now committed to choices that cannot be changed till their next inflection point is reached. 

The choice of mimik was driven by two related elements. In one camp there was the need to hone up the technical side of the offering - improving interoperability, security and the ability optimise the services delivered to the customers – and their customers’ customers. In particular there was the need to enhance the flow of data between the platform and the users’ own environment while at the same time also improving security, for the obvious reason that the data in transit carried some pretty sensitive personal data. This includes the use of two factor authentication. It was also important that it was possible to optimise the services in operation so that users could always have the chance to improve the profitability of each of their clients.

The other camp was a more direct interest to H2 - the ability to reduce its own operational costs as a by-product of such optimisation and interoperability improvements. This optimisation in particular meant that processing tasks that had been undertaken in the cloud could now be  undertaken on small systems directly in the customers’ own environment, a classic edge computing model. 

The cloud service H2 was using – and continues to use – is Amazon, and the cost for running the workload for one of its clients was $28,000 per month. By optimising the processes and moving to mimik-based systems as part of the client environment, the monthly Amazon bill has dropped to $2,800. Fakki notes: 

mimik said to me it would be probably close to half the cost, so we aimed at $14,000. But they just kept optimising and brought it down to $2,800. So they under-promised, but they also were very guarded because they there's so many things that are not in In their control, so they made a safe bet by saying half.

In addition, measurable benefits include a reduction in personnel required to keep hosting environments running down to about a quarter of what it was, while the speed at which they can build something has been cut in half. He is also aware of other factors are available should the company need them, such as mimik’s latency benefits, though these are not of immediate benefit to H2 at present. 

Logging everything, and everything else

One use case that would map onto a classic IoT-oriented edge computing environment would be a gym where at the moment every training system would be connected directly to an H2 instance running on Amazon. Using mimik, it becomes possible to equip each system with a low cost system that connects to the relevant sensors to log all relevant data about a user’s training needs, current performance, plus over- and under-achievements. That data can be then used not just by the gym but also other healthcare professionals and services that predict and plan future advice, activities and necessary healthcare interventions. Only final results and recommendations then need to be forwarded to the cloud services.

Fakki sees mimik playing a part in the company’s future plans, which involve what he refers to as ` “getting deeper into healthcare” with an extension of its work in health outcome tracking, and particularly patient-reported health outcomes. It has partnered with a Los Angeles-based startup Outcome MD and together they are targeting a single solution designed to provide medical practices with the tools to share, exploit and deliver health outcome insights across the spectrum of healthcare services providers.

This will work in both directions: down to the specific medical and health requirements of individual patients and their immediate healthcare providers, and upwards to the organisations handling the management of  their relationships with medical service and resources provision, in areas such as informing treatment decisions, strengthening medical documentation, revolutionising the visualization of patient care, delivering and applying best practices and enabling outcome-based care management that can include accounting for confounding factors, amongst other capabilities. 

These will include business management factors such as negotiating better contracts with payers, equipping outcome-based marketing, and generating new business opportunities.


A grey colored placeholder image