ECH is a leading provider of independent living and aged care services in South Australia.
Supporting thousands of clients across the region through a broad range of home support, wellbeing and residential services, ECH is recognized for its person-centered approach and commitment to continuous innovation in care delivery.
The challenge: Reducing administrative burden to increase client-facing care
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’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
- Formatting and structure across care plans were inconsistent
- Key well-being information was frequently omitted
- Staff were frequently diverted from client-facing care by this time-intensive process
- Documentation quality varied between care partners
The solution: Automation through GenAI
CGI delivered a seven-week proof-of-concept, designing and testing a generative AI (GenAI) 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 the following outcomes:
- Developed a GenAI-powered care plan generation system using AWS Bedrock and Claude Sonnet 3.5 v2
- Integrated with AWS Transcribe for audio transcription of wellbeing interviews
- Leveraged custom prompt engineering to create an empathetic AI Care Partner persona
- Implemented a secure AWS Landing Zone tailored for healthcare compliance
- Used 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 Australian dollars 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.
By leveraging generative AI, ECH has empowered Care Partners to focus more on person-centered care by significantly reducing the time spent on administrative tasks. Comprehensive and accurate care plans are now possible in just 36 seconds, freeing Care Partners to engage directly with clients and provide tailored, empathetic support.
Head of Digital Workplace Services, CIO, ECH Inc.
