Case Study
GreenShield was founded on the belief that all Canadians deserve access to better health, regardless of their circumstances. As Canada’s only national non-profit health care and insurance company, GreenShield pioneered the country’s first prepaid drug plan, and now leads with an innovative payer-provider model that integrates insurance and vital health services, simplifying care and improving outcomes for millions.
GreenShield launched its Robotic Process Automation (RPA) program back in 2020 supported by UiPath and Ashling. It quickly scaled with an employee crowdsourced and CoE led pipeline of automation opportunities. Within six months, they introduced AI and automation-powered enhancements to help Finance, Claims, and Provider Services teams streamline their workflows, reduce the burden of highly manual, repetitive tasks, and elevate the quality and purpose of their day-to-day work.
As automation matured, GreenShield began exploring more complex opportunities that combined AI with RPA—starting in the Contact Center.
Like many service organizations, GreenShield’s contact center was struggling to keep pace with rising email volumes. Each year, more than 500,000 emails poured in from members seeking care, resolving claims, or looking for answers about their health coverage.
Behind the scenes, agents were left sorting through overflowing work queues with little visibility into what was coming in, how often, or what needed urgent attention. There was no easy way to understand the types of requests being made, let alone how to prioritize them.
According to Andrew Kemble, Manager of AI Enablement and Delivery at GreenShield, volumes are expected to grow by another 20% annually. Kemble knew it was time to chart a smarter path, one that minimized operational clutter, cut through the noise, and allowed agents to focus on delivering meaningful support to members.
GreenShield partnered with Ashling to pilot UiPath’s Communications Mining (CM), a Natural Language Processing (NLP) and Machine Learning (ML) tool that interprets email messages, categorizes them, extracts key information, and analyzes sentiment at scale. Together, they launched a focused four-week proof of concept (PoC) in the Contact Center to explore how CM could uncover insights hidden within high volumes of unstructured email traffic.
Despite initial uncertainty, GreenShield’s openness to innovation, and trust in Ashling’s guidance led to a successful PoC. In just four weeks, GreenShield analyzed over 100,000 emails, accurately labeling and classifying incoming inquiries. For the first time, they had clear visibility into request types, patterns, and volumes. More importantly, the pilot revealed twenty business cases where AI and automation-powered improvements could enhance customer service quality.
With a successful PoC behind them, GreenShield and Ashling advanced to full-scale implementation. The teams evaluated the list of business cases using a value engineering framework, guided by input from GreenShield subject matter experts (SMEs).
From that evaluation, nine processes were prioritized. Ashling then designed, built, and deployed each solution using a multimodal approach that combined AI-powered tools with automation technologies such as Robotic Process Automation (RPA) and Intelligent Document Processing.
Each solution addressed common service interactions—such as identity verification—enabling agents to obtain consent quickly while safeguarding sensitive member information and privacy. Another enhancement automatically routes inquiries to the appropriate agent, minimizing delays and getting members connected with the right support as quickly as possible.
The result? Faster response times, quality care, and better experiences for both employees and customers.
Beyond email visibility, GreenShield uncovered insights that now inform agent training, service quality, and cross-functional decision-making. The team can isolate specific request types on demand to create targeted, real-world training for new agents. Sentiment analysis also reveals early signals of member response to product releases, feature rollouts, and business updates—providing measurable feedback for product and communications teams.
They also identified clear differences across channels: email inquiries tend to be more detailed, emotionally complex, and often used when members want something in writing, need clearer explanations, or a response they can reference later. These patterns now shape both training and communication strategies.
What started as a proof of concept became a model for how people and technology can work better together. By leveraging AI and automation, agents are empowered to focus where it matters most—on their members. It’s just one more way GreenShield is delivering on its mission of Better Health for All.
Technology Used
