From inbox to quote: AssurArt's digital transformation
How a Montreal-based insurance broker automated its most labor-intensive manual processes using generative AI.
AssurArt, a Montreal-based insurance broker specializing in the arts, technology, and environmental sectors, has embarked on a two-phase digital transformation with Mondrian, grounded in a clear customer-value approach. First came Hilma, designed to reduce friction in the acquisition process: automating the prospect’s journey, from the quote request to the delivery of documents. Next, Libero, to strengthen the long-term management of these customers: automating the sorting, classification, and archiving of emails in their CRM. Two distinct solutions, covering two critical moments in the customer lifecycle. AssurArt estimates that Libero’s ROI exceeds 300%, with more than 1,000 quote requests processed automatically via Hilma.
A niche brokerage firm that has chosen to transform itself from within.
Insurance Brokerage — Arts, Technology, Environment, Life Sciences
Boutique, ~10-25 employees
Canada
AssurArt is not your typical insurance broker. Specializing in demanding and often underserved sectors such as the arts and culture, the social economy, information technology, and the life sciences, the company has built a solid reputation through personalized service and rare industry expertise. To maintain this advantage, AssurArt identified two critical moments in its customer lifecycle: acquisition, where friction in the quote process could lead to the loss of prospects, and day-to-day management, where the volume of emails consumed valuable time that brokers would have been better off spending on advising clients. The transformation strategy was built around these two levers, which are closely tied to customer value.
Two challenges, two obstacles: customer acquisition and customer management.
AssurArt’s customer journey plays out on two distinct fronts. Upstream, winning over a prospect: a quote process that was too long or too complex was enough to lose them. Downstream, retaining existing customers: every incoming email had to be read, summarized, categorized, and archived manually in their Applied Epic CRM—a daily burden that ate into the time available for actual consulting work. Both friction points, at acquisition and at account management, were capping growth and service quality alike.
Manually sorting emails required each broker to handle up to 50 individual interactions per day. The risk of errors, misclassification, or oversights was constant, and even the slightest absence led to an unmanageable backlog. On the quote side, each request required several manual steps between the client form, the CRM, and the pricing platforms, slowing down response times and limiting the team’s ability to scale.
Three specific challenges made the situation difficult to resolve. First, Applied Epic, the market-leading brokerage software, offered limited API integration capabilities, forcing creative workarounds rather than a simple direct connection. Second, AssurArt’s workflows were highly specific to their niche industry, making generic off-the-shelf solutions unsuitable. Troisièmement, l'environnement IA évoluait si rapidement que tout investissement dans une solution trop rigide risquait d'être obsolète avant même d'être déployé.
First Hilma to acquire customers, then Libero to build loyalty: an approach that follows the customer lifecycle.
The deployment strategy followed the customer lifecycle: first, reduce friction at the entry point; then, optimize management over the long term. Hilma was designed primarily to address customer acquisition. Prospects fill out a short form on the AssurArt website, available in French and English for 11 types of coverage. In just a few seconds, Hilma automatically creates the lead in Epic, generates a pre-filled link to the corresponding long form, submits quotes to partner insurers, and sends summary PDF documents to the client and the sales team. Once the client is on board, Libero takes over for day-to-day management. This email processing engine runs in the background: every minute, it retrieves incoming emails, analyzes them using a language model, extracts a structured summary, identifies the corresponding client in Applied Epic, and automatically archives the interaction in the correct folder with the correct label. Brokers can then focus on what truly adds value: advice, relationships, and expertise.
The collaboration with AssurArt was structured around an iterative and pragmatic approach. Rather than waiting for a perfect solution, Mondrian and AssurArt chose to deliver quickly, learn through production, and gradually expand the system’s capabilities. The AssurArt team’s industry expertise—including their in-depth understanding of workflows, regulatory requirements, and their clients’ specific needs—was crucial in developing solutions that are truly tailored to their reality.
| Component | Role in the solution |
|---|---|
| LLM (OpenAI / Anthropic) | Analyzes, summarizes, and classifies incoming emails in Libero |
| FastAPI + Google Cloud Platform | Hilma's backend infrastructure, scalable and continuously deployed |
| Applied Epic API | Syncs with the brokerage CRM for prospect creation and activity archiving |
| Insly + Microsoft Graph | Automated quote rating and access to Outlook mailboxes |
Measured ROI, a relationship built to last
| Indicator | Result | Business impact |
|---|---|---|
| ROI on Libero | +300% estimated by AssurArt | Investment fully recovered within the first year of operation |
| Quote requests processed | 1,000+ via Hilma | Higher processing capacity with no added headcount |
| Emails handled by Libero | ~50 / broker / day, no manual work | Daily triage eliminated, time recovered for client advisory work |
| Partnership length | 2 years of continuous collaboration | Proof of a trusted relationship and lasting value, beyond a one-off project |
Facing a similar challenge in your industry?
A 30-minute call is enough to find out where AI can create value for you.
Planifier un Appel