AI challenge winners

We received an overwhelming number of entries for our AI challenge. Thanks to those who took the time to submit their incredible ideas. A big congratulations to the winners who contributed some fantastic concepts, you can read about the winning submissions below.

First place winners

Kate Cathcart: analysis of behaviour patterns to identify potential problems and improve consumer outcomes

Kate's submission recommends using AI to search case notes to identify patterns in consumer behaviour over time, and to access existing research to suggest remedies. 

When a consumer has a variety of support workers it is hard to recognise behavioural trends to determine what is a normal or expected response, or something that is uncharacteristic or a symptom of a problem.

Without regularity or someone reading every case note it would be easy to miss notable changes in behaviour. Recognising these patterns would not only prevent neglect when signs of a problem are present, but would also assist support coordinators and workers to make minor adjustments to improve consumer wellbeing. 

An example of how this process might work could be:

1. AI analyses adjectives in case notes and identifies the words disengaged, talkative and loud as the three most common words and determines that 'disengaged' is the outlier to the common adjectives for this consumer

2. AI analyses notes connected with the word disengaged, detecting that the word is 4x more likely to occur on Tuesdays

3. Using research or preprogrammed options, AI suggests for the support worker to ask some probing questions (e.g. If you could change one thing about this morning, what would it be?)

4. Combining case notes with these answers, AI could suggest that the consumer is going on long walks on Monday shifts as it is one of their planned goals, but is often sore and tired on Tuesdays

5. AI can use research to suggest some remedial options e.g. shorter more frequent walks, ensuring the consumer is hydrated or to plan for a restful day on Tuesdays

6. The support coordinator assesses the issue and suggested improvement and adjusts consumer alerts to notify support workers of the resolutions

7. AI reviews continuously and provides a comparison of present day to past highlighted concerns (e.g. 3 months later disengaged is only 2x more likely to occur on Tuesdays) 

Shannen Williams: supportability bot

Shannen's idea involves using a supportability chatbot allowing consumers to text a number instead of using an app.

Using a text messaging system would be ideal as consumers will feel as though they are texting a friend. It would make for a more personal experience and is far less confronting than using other services.

There would be no limit to what the supportability bot could do. For example, consumers with limited knowledge around their diagnosis and disabilities could use the supportability bot to find out more information about their particular needs (such as dietary requirements).

They could take a picture of what’s in their fridge/pantry and send it to the supportability bot and the bot could send related recipes and what ingredients would be needed to make the meals. It could also assist with money management when shopping, a consumer could message the supportability bot with something like “I have $40 and I need to make 3 meals from it.”

Consumers could also take photos of perishable foods (such as meat and dairy) and the supportability bot could remind them of when it’s about to expire.

It could also suggest the order in which ingredients should be used to minimise food wastage.

This would be such a powerful tool for our consumers, especially those who are in a vulnerable state, at risk of loneliness and those who struggle with their memory.

Because of the supportability bot’s friendly and non-confronting interface, it could also be used for consumers who are at risk of suicide, connecting them with supports and resources in the community.

We could also use the bot to contact consumers with reminders about their shifts for the day.

We could also include suggestions on what activities they could do or remind them of their goals and make suggestions based off these goals.

This would also assist with the regions that no longer have CSO’s.

First runners up

Nicole Guzowski | AI for support coordination

It is very important for support coordinators to utilise their time efficiently.

Nicole says AI could greatly assist coordinators in planning budgets, including linking services according to what is in these budgets.

AI could also be used to reference information from an intake form and crossreference against a consumer’s budget to create a list of local supports (within that budget) catering for the consumer's individual needs.

"clerical time could also be cut in half by having simple functions to document the work they do with their consumers," said Nicole. 

Crystal Webster | AI-powered fundraising optimisation

Crystal's submission involves the development and deployment of AI algorithms to optimise fundraising efforts.

By leveraging advanced data analytics Crystal believes we could analyse donor data, behaviour patterns and campaign performance to enhance fundraising strategies.

This initiative will enable us to identify potential donors, personalise fundraising appeals and determine the most effective fundraising approaches based on historical data and predictive analytics.

By maximising our fundraising efforts, Crystal anticipates a significant increase in donations and revenue, enhancing our financial sustainability and ability to serve our community.

Second runners up

Jemma Gultzow | Quick access to relevant policies and procedures

Jemma believes we can utilise AI to quickly access policies and procedures relating to a question or statement. 

AI could return with both written and visual instructions, or direct to a suitable department if unsure. For example, employees could ask AI about how to claim kilometres, how to apply for leave or where specific documents are located. 

This process could assist new employees and take some weight off our team leaders. 

Rebecca Sheeley | Assistance in navigating online programs (SharePoint, Aurion)

Rebecca's submission recommends assisting employees who encounter difficulties when navigating programs such as SharePoint, because locating the precise terminology for searching policies, procedures, forms or documents can be somewhat frustrating.

This has the potential to serve as a valuable resource for employees.

AI could function as a virtual help desk to provide information on various topics including SharePoint, Aurion, HR inquiries, internal documents, policies and procedures, career development, training opportunities, selectability values, careers and more.