One AI Agent for a Multi-Campus Education Group in Canada

One agent for the group works if four layers stay separate

A single AI agent can serve several institutions in a Canadian education group, provided the set-up keeps four layers apart: brand, knowledge base, routing and personal information. Without that separation, an applicant to one college gets an answer written for another, and the recruitment team on one campus sees contacts that belong to a sister institution.

Canada adds complications that a single-country template misses. Education is a provincial responsibility, so programs, funding and credential frameworks differ between provinces. Privacy law differs too, and Quebec has its own regime. The practical question is therefore not "one chatbot or several?" but "what is shared across the group, what belongs to each institution or province, and who decides?" This article offers a grid for answering it and complements our guide to AI agents for student recruitment.

A note on method: paid search-volume and difficulty data were not available for this topic, so the analysis rests on public research. The article contains no performance figures, and nothing in it comes from Skolbot customers.

Three architectures to choose between

Most groups want the third option because it matches how applicants think. Someone looking at a college diploma is not comparing it with a university degree in the same group, and an applicant in Ontario and one in British Columbia follow different application routes. The third option does require writing down who owns what.

Brand: one voice per institution

Each institution keeps its name, tone, domain and program list. The agent on its site introduces itself under that brand and discusses only its programs.

The group sets a narrower set of common rules than a style guide: what the agent never promises (an offer, a scholarship, transfer credit), how it discloses that it is an AI, and when it hands over to a person. If any institution serves French-speaking applicants, the French experience must be as carefully reviewed as the English one rather than machine-translated and left.

Decide in advance how the agent handles comparisons with sister institutions. Pointing to the sister institution when it fits better, or staying within its own remit, are both defensible. Improvising is not.

Knowledge: a group layer and an institution layer

Split content in two. The group layer holds what is truly shared: values, locations, general student support, legal notices, the privacy notice. The institution layer holds everything else: programs, tuition (often different for domestic and international students), application routes and deadlines, admission requirements, and recognition or accreditation.

A simple rule: any fact that differs by institution or province never goes in the group layer. Application routes illustrate it. Applicants to Ontario universities typically use a central application service, while other provinces and many colleges have different arrangements; confirm the route per institution rather than generalising.

Every item needs a named owner, a last-reviewed date and an expiry rule. We cover the method in keeping an admissions agent's knowledge base current. In a group, the extra risk is propagation: a correction made at one institution leaking into another, or never reaching the one that needed it.

Routing: getting each conversation to the right team

Routing answers two questions: which institution does this conversation belong to, and who picks it up? The most reliable signal is the entry point. A widget on an institution's site ties the conversation to that institution without asking the visitor to choose.

Ambiguous cases do arise: a group-level page, a fair, a link shared on social media, a visitor weighing two programs or two provinces. The agent then asks a sorting question (level, field, campus, residency status such as domestic, out-of-province or international) before directing. If the answer stays unclear, it offers a conversation with an advisor rather than guessing.

Plan routing inside each institution too: by campus, program, language or study mode. Each destination needs a team, response hours and a system where follow-up is tracked.

Personal information: who decides, who processes, who sees what

The federal Personal Information Protection and Electronic Documents Act (PIPEDA) applies, in the Privacy Commissioner's words, to "private-sector organizations across Canada that collect, use, or disclose personal information". Some provinces have their own private-sector laws that apply instead within the province, and Quebec's is the best known. Public colleges and universities are often subject to provincial public-sector or freedom-of-information and privacy statutes instead. Which regime applies to each member of the group is a legal question; get a written answer before launch.

For Quebec members, see our parallel article for Quebec and the Commission d'accès à l'information on Law 25. If any member sends personal information outside its province or the country, ask the vendor where data is hosted and supported.

Roles follow who decides the purposes and means of processing. If each institution is a separate legal entity recruiting for itself, it is normally responsible for its own applicants' information. If the group centralises decisions (one tool, one shared database, shared campaigns), write down the split of responsibility. The vendor processes information on the institution's behalf, and the contract should cover instructions, security, hosting location, sub-processors and what happens to the data at the end.

Three working rules follow:

1. Separation by default. An advisor at institution A sees only institution A's conversations and contacts.
2. No silent reuse. An applicant who inquired at one institution is not contacted by a sister institution without valid consent or another valid basis, and clear notice.
3. Aggregated group view. Group leadership tracks aggregated indicators without access to named records unless there is a defined, documented need.

Governance: who changes what, and how

Then define a change path: who proposes, who approves, who publishes, how to roll back. Changes to the group layer go through the platform owner plus a lead from each affected institution. Changes to an institution layer go through that institution's lead alone. Keep version history, so that when an applicant disputes an answer you can see what the agent knew that day, and review handed-off conversations regularly, institution by institution.

Measuring without mixing institutions

Each institution has its own funnel, seasons and brand recognition. Comparing two uncorrected creates false internal rankings. Measure each against itself, over comparable periods, and consolidate at group level only operating indicators such as response time, handoff rate and content flagged as missing. Our evaluation grid for choosing an AI agent lists criteria to test with a vendor, including multi-site handling.

Questions to put to a vendor

• Can separate brands, domains and knowledge bases be configured in one workspace, in English and French?
• How are access rights managed by institution and by role?
• Where is data hosted, and what are the terms for sub-processors?
• How are an institution's data exported and deleted if it leaves the group?
• Which CRM integrations exist for the systems each institution already uses?

Take only what the vendor shows working.