Case study

Isobloc’s president built AI agents tailored to Quebec’s construction industry

In Mascouche, Eric Dionne developed five artificial intelligence agents to support his company. With Agence 3A, he then established a methodology and roadmap to further develop these tools and prepare for broader use.

The Isobloc logo over a building under construction

Preparing a quote for masonry blocks requires calculating quantities, accounting for corner blocks, and factoring in pallets and accessories. At Isobloc, these business rules now guide an AI agent that prepares quotes in the company’s management system. Eric Dionne reviews each quote before it is sent.

The estimator is one of five agents the president built using Anthropic’s Claude. From identifying business opportunities to forecasting cash flow, he has gradually integrated AI into several functions at his company, which specializes in insulated masonry blocks designed and manufactured in Quebec.

“If I hadn’t done this work with artificial intelligence, I wouldn’t be where I am today.”

Eric Dionne, President of Isobloc

The founder’s expertise at the heart of the tools

Eric developed his agents one at a time, based on Isobloc’s needs. Each handles a specific activity and draws on the company’s information and ways of working.

The tender monitoring agent identifies projects and companies to approach, then feeds that information into the CRM. The prospecting agent reviews business opportunities and drafts personalized follow-ups. The estimator applies calculation rules specific to Isobloc’s products. Another agent helps prepare documents to protect the company’s payment rights and tracks project-related deadlines.

The virtual CFO works with a 13-week cash flow horizon. It prepares projections and analysis to support the president’s decisions. Eric remains responsible for the decisions and reviews built into these processes.

Results in day-to-day operations

Eric estimates annual savings of more than CAD $60,000 across commercial activities since implementing his ecosystem. He also reports preparing cash flow forecasts in a single evening, an exercise that previously stretched over a month.

The work also extends to the customer journey. Isobloc redesigned its website and brought its product data sheets, installation guides, and design documents together in a technical library for construction professionals. The website, incoming inquiries, and CRM all contribute to the same business development effort.

An internal inventory dated July 15, 2026, recorded nearly 2,500 document access permissions granted for this library, more than 2,000 of which had been used at least once. These permissions are counted separately for each document and person.

According to Eric, many construction companies could benefit from AI to increase productivity and reduce staffing levels. His experience at Isobloc has shaped this conviction: better integration of technology can change the resources a company needs to operate.

Isobloc plans its next stage with Agence 3A

Eric engaged Agence 3A to assess the tools he had built and determine how to develop them further. Agence 3A’s technical team, which includes Anthropic-certified developers and architects, mapped the five agents, analyzed how they worked, and identified development and governance priorities.

The engagement resulted in an assessment, technical recommendations, and a roadmap.

Agence 3A’s contribution connects the applications envisioned by the business leader with a structured development methodology: organizing knowledge by function, documenting processes, defining usage rules, and planning connections to the company’s systems. This support gives Eric a framework for choosing the next improvements and preparing for their deployment.

The approach illustrates Agence 3A’s offering for small and medium-sized businesses: start with existing needs and tools, establish priorities, then support AI integration and governance. The roadmap also includes measuring gains to direct investment toward the applications that deliver the most value to the business.

Eric remains committed to being directly involved in designing his tools. Agence 3A therefore proposed a collaborative approach that allows him to develop his skills as the project progresses. The next stage combines his industry knowledge with Agence 3A’s technical expertise.

“Agence 3A helped me rethink the foundation of my AI ecosystem and improve communication between my agents and my data source. This will make it easier to scale the agents during Isobloc’s next phase of growth.”

Eric Dionne, President of Isobloc

Common questions

How can a construction business use AI agents?
The Isobloc case illustrates five uses: monitoring tenders, supporting prospecting, preparing quotes, preparing certain documents and producing cash flow forecasts. The agents use the company’s data and business rules. Isobloc’s president, Eric Dionne, remains responsible for decisions and the reviews built into these workflows.
Can AI help prepare construction quotes?
Yes. In the Isobloc case, an agent prepares quotes using calculation rules specific to the company’s products, including quantities and accessories. Eric Dionne reviews each quote before it is sent. The case illustrates AI-assisted preparation with human review based on the company’s business rules.
What benefits does Isobloc report from using its AI agents?
Eric Dionne estimates annual savings of more than CAD $60,000 across commercial activities since implementing his AI ecosystem. He also reports preparing cash flow forecasts in one evening, compared with an exercise that previously stretched over a month. These gains reflect Isobloc’s specific uses and business context.
What was Agence 3A’s role in Isobloc’s AI project?
Eric Dionne had already built the five agents. Agence 3A then assessed the tools, analyzed how they worked and identified development and governance priorities. The engagement produced an assessment, technical recommendations and a roadmap to organize knowledge, document workflows and prepare future integrations.
Which decisions remain with the team when AI is used?
At Isobloc, Eric Dionne reviews quotes before they are sent and remains responsible for decisions supported by the analyses and forecasts. AI prepares information and supports the work, while review points remain part of the workflow. Other businesses need to define these responsibilities around their own activities and risks.

Let’s talk about your business.

Tell us what your team is working on and where you need support.