Age UK Islington
Zero Subjectivity in Care: Empowering Age UK Islington with Cognitive AI
Age UK Islington faces a profound challenge: an ageing population driving unprecedented demand for life-enhancing services. But scaling care isn't like scaling software. It’s deeply, wonderfully human.
<>
The Challenge: The Paradox of Perception
Volunteers are the lifeblood of the operation, undertaking the vital task of reaching out to the elderly to understand their circumstances and anticipate their needs. Yet, this reliance on human interaction presented a hidden flaw: subjectivity. One volunteer might hear stoic resilience; another might hear quiet desperation.
This inconsistency wasn’t just a process issue - it was a systemic risk. Relying solely on individual perception meant the level of support varied from caller to caller. At its worst, the elderly people who needed help the most might slip through the cracks. To scale effectively, Age UK Islington needed a unified, consistent gauge to ensure the right care reached the right people, every single time.
<>
The Response: Standardising Empathy
Razor was brought in to build a bridge between human intuition and data-driven certainty. Our mission? Deliver a technology prototype that transforms a subjective guessing game into a streamlined, standardised process - without losing the human touch.
Phase 1: The Guided Conversation
We started by securely extracting call information from their Microsoft Dynamics CRM. Instead of leaving volunteers with a blank slate, we engineered a guided, intelligent workflow.
At each step, volunteers ask a series of personalised, pertinent questions. This shift was fundamental: we stopped merely capturing what the volunteer thought they heard, and started capturing exactly what the caller said.
Phase 2: Cognitive Sentiment Analysis
Words alone only tell half the story; the emotion behind them reveals the truth. By integrating Microsoft’s Azure Cognitive Services, we applied advanced sentiment analysis to the conversation data. The system intelligently processes the text to determine the underlying emotional tone of the caller, turning qualitative feelings into quantitative insights.
Phase 3: The Augmented Baseline
We didn't replace the human; we supercharged them. By combining the volunteer’s subjective perception with the AI’s objective emotional data, we established a consistent baseline. For the first time, Age UK Islington had a reliable, unified method to determine whether an individual was coping or urgently needed assistance.
<>
The Result: Consistency Meets Care
The prototype is currently being trialled by Age UK Islington’s volunteers, and the early feedback has been overwhelmingly positive. But we aren't stopping there.
As more data flows in, we’re continually refining the delicate balance between human intuition and artificial intelligence. By introducing new metrics and predictive analysis, we are evolving the system into a perfectly tailored tool. Age UK Islington is no longer just scaling operations; they are scaling empathy.
Why Razor?
Because we believe in doing things "right." This project wasn't about ripping out a CRM or building a cold, robotic script. It was about applying bleeding-edge cognitive AI to solve a deeply human problem. We brought the enthusiasm, technical firepower, and patience needed to navigate the delicate intersection of technology and elderly care.
By introducing new metrics and predictive analysis, we are evolving the system into a perfectly tailored tool. Age UK Islington is no longer just scaling operations; they are scaling empathy.
Andy Murphy
Chief Executive Officer at Age UK Islington
Proactively contacting people – instead of waiting for them to contact us – is a vital part of our future services strategy. The reason it’s not usually done is because it’s difficult and involves large scale mobilisation of volunteers.
Razor have helped us work through these challenges with enthusiasm, creativity, incredible technical skills, and… patience.
We only work with people who want to do things ‘right’ – and they do.
