Healthcare
123 Dentiste
A Quebec dental network deflects a third of its call centre volume with an AI front desk.
32.5%
Call-centre volume deflected
1,600
Calls triaged per week
RetailPressure washingMontreal, QC
Powerplay Pro
A Montreal pressure washing company answers every support call with AI.
Devpro built a virtual customer support agent for Powerplay Pro, a Montreal-based pressure washing company, integrating it with Freshdesk and an Asterisk PBX to handle inbound calls, triage issues, create and update tickets, and send follow-up communications, all without staff touching routine support interactions.
30
Support calls handled daily
8
Tickets auto-logged per day
24/7
Coverage without added headcount
01The challenge
A growing service company fielding support calls with no system behind them.
Powerplay Pro was handling around 30 inbound support calls per day but had no structured workflow to log, track, or follow up on those interactions. Every ticket was created manually, every follow-up was done by hand, and no two calls were documented the same way.
- No ticket creation at call time. Support issues were handled verbally and logged later, or not at all. Repeat callers had no documented history to reference.
- Manual follow-up overhead. Status updates, appointment confirmations and troubleshooting instructions were sent by hand, pulling time away from field operations.
- No caller history on pickup. Agents had no way to identify whether an incoming caller had an open ticket, requiring customers to re-explain their issue from scratch on every call.
02The approach
Connect every call to a ticket, without touching a keyboard.
We scoped this as a full-stack integration from the phone system to the support platform. The agent needed to identify callers, retrieve their history, act on it, and follow up, all without staff involvement on routine contacts.
01Wk 1
Discovery
- Call flow mapping
- Issue type catalogue
- Freshdesk API review
02Wks 2–3
Build
- Agent conversation design
- Freshdesk API integration
- SMS & email module
03Wk 4
PBX integration
- Asterisk routing
- Call handling logic
- Fallback configuration
04Wks 5–6
Testing & deployment
- End-to-end testing
- Workflow validation
- Production deployment
03The solution
A virtual agent that knows the caller, logs the issue, and follows up automatically.
Every inbound call is identified, triaged, and documented in Freshdesk without staff intervention. Follow-up by SMS and email happens automatically based on call outcomes.
Caller ID ticket lookup
The agent identifies callers by their caller ID and automatically retrieves any open Freshdesk tickets, giving the conversation immediate context without the customer having to re-explain.
Automatic ticket creation
When no prior ticket exists, or when a new issue arises, the agent creates a structured Freshdesk ticket capturing the issue details, caller information, and diagnostic data collected during the call.
Ticket status updates
Existing tickets are updated in real time based on caller input and workflow logic, keeping Freshdesk accurate without a staff member touching the record.
SMS and email follow-up
Ticket confirmations, status updates and troubleshooting instructions are sent automatically to the customer by SMS or email at the conclusion of the call.
Asterisk PBX integration
All inbound calls route through the Asterisk PBX directly to the virtual agent. The integration handles call receipt, routing logic and fallback paths for calls the agent cannot resolve.
04Results
Thirty calls per day handled without staff involvement.
Measured since deployment, against the prior state where all support interactions required direct staff time for logging and follow-up.
01
30
Support calls handled by AI daily
Without staff involvement on routine contacts
02
8
Tickets auto-created in Freshdesk daily
No manual logging required
03
24/7
Support coverage without additional headcount
Calls outside business hours are automatically logged
04
0
Manual follow-up messages sent by staff
All confirmations, updates and instructions sent automatically
Measurement window·Post-launch steady state
Baseline·Manual support workflow pre-deployment
Sample·~30 calls / day average
