Introducing... CareTec.AI

CareTec.Ai is a AI/machine learning platform we built for the care sector. Our aim is to help care homes embed data analytics into their care management software.

Background

The care sector is facing a major problem. The Office of National Statistics (ONS) reports that for years to come, the UK population is ageing.

For over 65's, this process will continue so that by 2025 there will 25% increase in people needing care from 2015.

To that end, we adapted our AI platform to achieve a single goal. That is to improve health and social care by turning data into actionable insight. Our platform uses cutting-edge machine learning and AI techniques. 

It's the next step in data analysis for health and care. Through CareTec.AI  we:

  • Give care providers the insight they need to deliver care
  • Give technology providers the solution they need to turn their data into insight

Origins

We spoke to care software providers throughout the UK and Europe. We are also members of CASPA (Care Software Providers Association).

During that process, some of our clients said they wanted to get started straight away.

We are working with others to help them identify how CareTec.AI will add value to their product.

Requirements

Our clients need several things:
  • Guidance on preparing their data for CareTec.AI
  • Guidance on how to visualise the output from our training models
  • Help and support with using the data ingestion endpoints we provide

Solving these issues means technology providers integrate easily with our platform.

That means they have AI-powered analytics right at the heart of their software.

Approach

We use the CRISP-DM process. It stands for Cross-Industry Standard Process for Data Mining.

It provides a six-step process that looks like this:

  1. Business understanding
  2. Data understanding
  3. Data preparation
  4. Modelling
  5. Evaluation
  6. Deployment

It is cyclical. So we may iterate through steps 1 and 2 in particular, to ensure we have a solid understanding of what's needed. 

We'll work with you to prepare your data for analysis. Plain old CSV files work for us. Or you can send us data via our API endpoints.

In the modelling phase, we run your data through our machine learning algorithms. They are purpose built to work with any time series data.

Also, we don't need large data sets. It's fine if that's what you have. But there is a common misconception that AI and ML only work on _big data_. That's not the case with our algorithms.

Results

While our running projects are quite new, we have already seen results.

For example, on running some fluid intake data for residents in one care home, our algorithms reported an anomaly. It turned out that the resident was suffering with an infection.  

For another client, we have shown them how a multi-step care record form is still useful for data analysis. With the right data munging to optimise storage. 

Benefits

So far, our clients have:

  • Been alerted about anomalies in care record data indicative of ill health. That's before symptoms were clear
  • Seen how cleaning complex data works
  • Learned how AI and ML provide actionable insights

Next steps

CareTec.AI works, we know that and so do our clients. The next step is to create realtime alerts. That way, carers will know when there is a non-normal pattern occurring for a given resident.

We are working with our clients to assess the best way to role out triggers and alerts. The preferred option is through their care management systems dashboards.