
A Quebec waste management company replaces hours of receptionist intake with two bilingual AI agents.
Ricova, a Quebec waste management company, replaced their receptionist-handled complaint and dispatch intake with two dedicated bilingual AI voice agents integrated directly into their Webex phone system. Each agent replicates the full receptionist conversation, captures the request, and fires an email to the appropriate team — handling 185 calls and 400 minutes of conversation per month.
A receptionist spending hours on intake calls that followed the same script every time.
Ricova's inbound call volume for customer complaints and dispatch requests landed entirely on the receptionist. Every call required the same structured conversation — gathering details, clarifying the request, identifying the right team — and then routing the information manually to whoever needed to act on it. The process was reliable but expensive in time, unscalable, and left requests sitting until someone processed them.
- Routine intake consumed hours of receptionist time. Complaint and dispatch calls follow a predictable pattern: same questions, same information needed, same teams to notify. But every call still required a person on the line to run the conversation from start to finish.
- Information handoff was manual and delayed. After each call the receptionist had to relay the captured details to the appropriate team — by email, by message, or by walking the information over. That step introduced lag between when a request came in and when it was acted on.
- Bilingual demand added coordination overhead. Ricova's client base spans French and English speakers. Managing language routing across a manual intake process added another layer of coordination on top of an already repetitive workflow.
One agent per intake type. Both living inside Webex.
We scoped two purpose-built agents — one for complaint intake, one for dispatch requests — rather than a single combined flow. Separating them kept each conversation focused and allowed the post-call email to route precisely to the team responsible for that request type. Both agents were built to live natively within Webex so Ricova's team could manage them like any other extension.
- Receptionist call flow audit
- Complaint vs. dispatch intake mapping
- Webex extension structure review
- FR + EN complaint agent
- FR + EN dispatch agent
- Post-call email routing per agent
- Webex transfer configuration
- Webex IVR integration as extensions
- Live call testing
- Email delivery validation
The same conversation the receptionist had. Completed automatically. Emailed instantly.
Each agent runs the full intake conversation — the same questions, the same structure, the same information gathered — and sends a formatted email to the right team the moment the call ends. Both agents live in Webex as extensions, meaning they fit into the existing IVR and can transfer to any person in the organization without the caller ever leaving the phone system they already know.
A complaints agent and a dispatch agent each handle their respective call type end-to-end. Separating the flows keeps every conversation focused and ensures the post-call email reaches exactly the right team without manual triage.
Both agents detect language on the first exchange and conduct the full conversation in French or English without requiring the caller to choose. Ricova's full client base is covered without routing or staffing adjustments.
The agents live inside Ricova's existing Webex environment as extensions. They appear in the IVR menu like any other option and can transfer calls to any person in the organization — no parallel system, no external handoff.
The moment a call ends, the agent fires a structured email to the team responsible for that request type. The team receives the full context — caller information, issue details, request specifics — without waiting for the receptionist to relay it.
Requests are captured in real time during the call and dispatched immediately on completion. There is no lag between intake and action — the team receives the request the instant the caller hangs up.
Four hundred minutes of intake handled automatically. Requests dispatched the moment the call ends.
Measured across steady-state monthly volumes since deployment of both agents.
More work
All case studies

A Quebec driving school generates a 400% ROI from a targeted outbound AI campaign.
