maica.ai - Artificial Intelligence for Healthcare Providers

The Challenge Statement

The challenge in the NDIS space is counter-intuitive. The point of uniqueness of healthcare service providers is to offer more and more personalised participant experience and choice. Whilst this fosters well-being and a high level of service, it is also not sustainable due to human capital expenses, shortage of digital solutions, and a capped ability to charge for services. We believe that it is this trajectory of ever-increasing personalisation will ultimately make NDIS service providers unsustainable and therefore actually reduce the availability, quality, and responsiveness of service delivery.

Simply put, the cost of managing care (treatment and resource planning) is high due to it being a highly manual process. This high cost is causing providers to be unsustainable as funding regimes change, and the cost of servicing increases due to tight labour markets.

The Solution Statement

The solution to this challenge is a strategic thought change (in many ways counter-intuitive) to automate and simplify the choices provided to participants, standardise treatment and resource plans (and deal with exceptions to this via human capital), as well as use historical medical data and computing power to commoditise large parts of the lifecycle.

The solution to this challenge is based on a set of assumptions, including:

• The majority of treatment plans will follow a similar path to either health or wellbeing outcomes

• NDIS service providers are motivated to increase utilisation and resource optimisation of their team

• Historical data can accurately predict (based on sufficiently documented needs) future outcomes

• Future funding models will create more constraints on provider profits as sovereign risk tries to curtail spiralling government spending

• Providers will have to seek new funding streams such as “Fee For Service” to maintain sustainability

• This solution will have portability into the Ageing sector as it too is transformed.

We dynamically match needs to workforce to budget and are shifting the use of almost unlimited computing power in the health sector from operational to strategic to empower greater community wellbeing.

What is maica.ai

maica.ai is a digital platform that uses machine learning and artificial intelligence to accurately predict participant’s treatment plan and resourcing requirements as well as intelligently schedules all available human resources. This will reduce the operational human capital burden (which is cost-incurring) for NDIS service providers, allowing them to focus on delivering services (which is revenue-generating) by using structured historical data and the enormous computing power available today to increase efficiency and therefore providing the participant better healthcare services.

In nutshell, maica.ai will perform the following functions:

• Provide a digital platform to manage participant’s profiles (including personal, medical, background, and engagement information)

  • An optional connection to the Salesforce macia.me solution is available to integrate with an enterprise-level CRM for larger organisations.

• Provide a digital platform to manage a health worker’s profile (skills, location, demographics) so that these attributes can be matched to a participants needs.

• Structured Needs Assessment Tools to properly understand a participant’s challenges and is integrated with existing third party systems that generate NAT’s.

• An AI and ML-driven platform to determine the most likely treatment & resource plan to succeed, for example, a participant may be recommended the following treatment plan (auto-derived by maica.ai):

  • 4 sessions of Speech Therapy will lead to 85% better outcomes whereas 10 sessions will lead to 90% better outcomes, therefore the recommendation is: 4 Sessions of Speech Therapy

  • 3 sessions of Occupational Therapy with a Level 3 Care Worker will lead to 87% recovery whereas 5 sessions of Occupational Therapy with a Level 2 Care Worker will lead to 79% recovery, therefore the recommendation is: 3 Sessions of Occupational Therapy with a Level 3 Care Worker

• An AI-driven resourcing and scheduling engine to intelligently allocate the most appropriate human resources (health workers) based on a set on inputs, including:

  • Plan Budgets and applicable services

  • Resource Availability, Skills, & Conflicts

  • Participant Preferences (Limited)

  • Geographic Constraints (Routing/Traffic)

• An omni-channel platform for participants to self-manage the following:

  • Personal Details

  • Plan Details/Statements/Invoice History

  • Service Times/Resources (this will be constrained to allow the AI-driven algorithms to continually optimise the workforce (health workers), an example is provided below:

    • A participant may currently have a service scheduled for Thursday at 3pm. This time is no longer suitable, so rather than being able to suggest/request a new time, maica.ai will present up to 3 options to select from which fall within the optimised workforce management plan. The participant will then be able to select the time suitable to their needs.

    • A participant may currently have a service scheduled with a particular care worker with whom they do not wish to work with anymore. The maica.ai platform will, similar to the above, propose a number of available care workers who are suitable for the work to be performed, are scheduled to be in the geographic area at the scheduled time, and meet all the above-mentioned constraints also.

  • A historical medical database that is fed (participant’s to be de-identified) by medical practitioners, researchers, and other relevant parties to provide the underlying data of success/failure of particular treatment strategies and their outcomes. This “care plan design team” will utilise best practice clinical evidence and their clinical judgement to derive a standardised care plan for particular input parameters. Of course unique conditions and outliers will exist, but the team will build standardised offerings that are holistic across clinical and non-clinical care, and this will be correlated to available funding to build the plan. If funding is not available, it may be possible that our “care plan design team” can support participants in advocating for better funding.

  • Any exceptions to the above process, which is underpinned by AI-driven platforms, will be handled via existing human processes. This will continue to exist but will be dramatically reduced by commoditising the majority use case.

• Automatic Invoice generation, claiming, and payment engine which allows healthcare providers to register their bank account details for fast payments within hours of submitting an Invoice. We anticipate that this function will ultimately be replaced by the NDIS payment system (currently under RFP) but some healthcare providers may elect to remain on our solution regardless.

• A published API which will accept a set of structured parameters and return a treatment/resource plan as well as schedule. This uses a machine-learning model similar to https://docs.aws.amazon.com/machine-learning/latest/dg/tutorial.html by using Amazon’s ML technology platforms (very likely but design yet to be confirmed).

How are we building maica.ai

To be completed

Why are we building maica.ai

The current healthcare provider market is struggling to provide care services at a level that is financially sustainable whilst also giving participants freedom of choice to create their own wellbeing journeys. With increasing regulatory reform that stemmed from the Royal Commission and now funding reform (AN-ACC and Home Support Program) that are due in the next 12 months, providers will no longer be able to rely on government block funding to sustain their finances. We expect providers to be squeezed out unless they can improve efficiencies and remove costs that are unlikely to be directly funded activities.

It is our belief that by making healthcare providers more efficient, we can ultimately contribute to the greater wellbeing of the people who need it whilst also building a commercially profitable platform offering the following key services:

• Participant Management

• AI-Driven Treatment and Resource Plans (based on historical medical data)

• AI-Driven Scheduling and Rostering (based on limiting participant choice)

• Integration into the macia.me Salesforce solution to cater for the Enterprise

The Commercial Model

The commercial model can be found here and is based on three separate revenue streams, including:

• Cost per managed Participant

• Cost per processed Invoice

• Cost per Bank Account Transfer (for Rapid Payments)