Tailored client care plans were proving time intensive, inconsistent and often incomplete. ECH needed a solution to improve the quality of the care plans and free up staff to provide more client-facing care.

ECH is a leading provider of independent living and aged care services in South Australia. With approximately 840 employees, ECH supports thousands of clients across the region through a broad range of home support, wellbeing and residential services.

The organisation is recognised for its person-centred approach and commitment to continuous innovation in care delivery.

The challenge

ECH’s Care Partners spent up to 60 minutes per client manually compiling care plans based on multiple well-being assessments. This process presented several challenges:

  • Only 9.8–78.1% of information from source documents was captured
  • Inconsistent formatting and structure across care plans
  • Frequent omissions of key wellbeing information
  • This time-intensive process diverted staff from client-facing care
  • High variability in documentation quality between Care Partners

The solution - automation through GenAI

CGI delivered a seven-week proof of concept designing and testing a generative AI solution for care plan automation. The AI system transformed audio and document-based assessments into comprehensive, structured care plans aligned to ECH standards - automatically and in seconds.

The work included:

  • Development of a Gen AI-powered care plan generation system using AWS Bedrock and Claude Sonnet 3.5 v2
  • Integration with AWS Transcribe for audio transcription of wellbeing interviews
  • Custom prompt engineering to create an empathetic AI Care Partner persona
  • Implementation of a secure AWS Landing Zone tailored for healthcare compliance
  • Use of AWS Lambda for serverless orchestration and automation

The solution delivered significant improvements in efficiency, accuracy and care quality. Care plan creation dropped from 20 minutes to just 36 seconds, saving over 3,000 hours annually across 2,300 clients. Accuracy rose from 35.3% to 92.5%, while consistency in content and formatting exceeded 85%.

These gains are projected to save $138,060 per year. The quality of care also improved through more comprehensive plans, better data capture, reduced risk of omissions, stronger compliance and more time for client interaction.

Following the success of the proof of concept, ECH proceeded with a production implementation of the solution, signaling confidence in the long-term benefits of generative AI in care delivery.

The project referenced in this case study was delivered by BJSS, which CGI acquired in February 2025. 

3,297%
efficiency gained in care plan generation time – 20 minutes to 36.4 seconds.
 
 
3,068 hours 
saved annually across 2,300 clients
By leveraging generative AI, ECH has empowered Care Partners to focus more on person-centred care by significantly reducing the time spent on administrative tasks. Comprehensive and accurate care plans are now possible in just 36 seconds, freeing up Care Partners to engage directly with clients and provide tailored empathetic support

Sharon Paulson Head of Digital Workplace Services, CIO, ECH Inc.