Tourism · Mirstravel · travel agency
Mirstravel: an AI agent that answers, quotes, and hands off travel inquiries on WhatsApp
The agency answers on WhatsApp around the clock: the agent covers destinations, dates, and financing, builds the quote, and hands off to the advisor with the full record.
- Client
- Mirstravel · travel agency
- Sector
- Tourism
- Product
- Mirstravel
In numbers
Cost model
Behind the number: 4x sales capacity with the same team
Data provenance. Capacity (4.00x), sales time (−75%) and equivalent operating cost (48%): KPIs validated by the client. Volumes, rates, costs and the revenue effect: model rebuilt by Possition.
- Upfront investment
- $11,380
- Net annual benefit
- $12,984
- Payback
- 11 months
- 3-year NPV at 10%
- $20,909
- 3-year ROI
- 242%
See the full cost breakdownAssumptions, baseline, investment, run cost and benefit
Assumptions
| Inquiries per month | 1,200 on WhatsApp |
|---|---|
| Advisor time per inquiry | 12 min before → 3 min after (−75%) |
| Advisor loaded cost | $9.40/h (client's market) |
| After-hours inquiries | 38% of total |
| Inquiry-to-booking conversion | 3.0% → 3.5% (modeled) |
| Average margin per booking | $180 |
| Discount rate / horizon | 10% per year / 3 years |
| Item | USD |
|---|---|
| Advisor hours on answering, quoting and qualifying240 h/month × $9.40 × 12 | $27,072 |
| Total per year | $27,072 |
| Item | USD |
|---|---|
| Discovery sprint (fixed fee) | $1,500 |
| Agent, portal and broadcasts: 190 h × $52 | $9,880 |
| Total one-time | $11,380 |
| Item | USD |
|---|---|
| Language models | $1,680 |
| WhatsApp Business API | $1,140 |
| Portal hosting | $900 |
| Managed service (catalog, evaluations, tuning) | $3,600 |
| Total per year | $7,320 |
| Item | USD |
|---|---|
| Hours freed: 180 h/month × $9.40 × 12 | $20,304 |
| Total per year | $20,304 |
Opportunity cost: the alternatives we evaluated
Before building, we priced every reasonable way to solve the problem, including doing nothing. The chosen option has to beat the best alternative, not just the status quo.
| Option | Upfront | Annual cost | 3-year net value, cost only (higher is better) | Assessment |
|---|---|---|---|---|
| Hire a third advisor | $0 | $18,000 | $0 | Adds 1.5x capacity at its own cost and doesn't cover nights or weekends. |
| Decision-tree chatbot (SaaS) | $800 | $1,188 | $20,002 | Deflects FAQs (≈30% of the time) but can't quote from the live catalog. |
| AI agent with portalChosen | $11,380 | $7,320 | $27,572 | Answers, quotes and qualifies 24/7; the advisor steps in only on real intent. |
Opportunity cost
The best alternative ruled out was the SaaS chatbot, with a 3-year net value of $20,002: that is the opportunity cost. On cost alone, the agent generates $27,572 ($7,570 more). Adding the modeled revenue effect ($12,960/yr from answering after-hours inquiries in minutes), 3-year value rises to $66,452.
Decision
$11,380 invested and $12,984 in net annual benefit on cost alone: payback in 11 months (5 months including the revenue effect) and a 3-year NPV of $20,909. Net savings equal 48% of the subprocess cost, the same figure the client reports.
Revenue effect (modeled, not part of the validated KPI): 38% of inquiries arrive after hours. Answering them in minutes lifts conversion from 3.0% to 3.5%: 6 more bookings a month × $180 = $12,960 a year.
US reference: with a travel agent at a $31/h loaded cost, the 2,160 hours freed per year are worth $66,960.
Headline metrics are validated by the client. The cost model is a reconstruction by Possition using the assumptions shown; it is illustrative, not an accounting audit. Run your own numbers in the ROI calculator, or walk through them with us in a discovery call.
The system in production
Real screenshots of the live system. Customers' personal data is hidden.
Mirstravel advisor handoff panel with destination, departure date, passengers, and quote for each inquiry 
Mirstravel broadcast scheduler with a WhatsApp message preview and sending window 
Mirstravel portal on a phone: tools, advisor handoffs, and message scheduler
The initial problem
Repetitive queries arriving at all hours, answers that depend on the salesperson's availability, and opportunities that go cold before anyone replies.
The previous process
Each inquiry was answered by hand, repeating the same information on destinations, departures, and payment options, with no organized record of who asked what or where it stood.
The solution implemented
An AI agent on the agency's WhatsApp that answers questions about trips, destinations, and financing using the up-to-date catalog (hotel, seats, dates, and boarding points), builds the quote, and hands off to an advisor only when there is real intent. Everything lives in a dedicated portal: a conversation log exportable to Excel, a handoff panel with destination, departure date, passengers, quote, and status (pending, urgent, handled), and a WhatsApp broadcast scheduler with a contact list and a sending window.
How we did it
Mapping frequent inquiries
We analyzed what people actually ask and what the advisor needs in order to quote, so the agent covers that ground completely.
Catalog and quoting
We connected the agent to the trip catalog with departures, hotels, and seats, so it answers and quotes with current information.
Handoff to the advisor
We defined with the team what turns an inquiry into an opportunity. The agent sends it to the panel with destination, date, passengers, quote, and the full conversation.
Broadcasts and follow-up
We added a message scheduler for departure, payment, or pending-document reminders, with a contact list and staggered sending.
Integrated systems
- WhatsApp Business
- Trip catalog
- Handoff portal
- Conversation log (Excel)
- Broadcast scheduler
The result
Immediate service at any hour, less team time on repetitive queries, and a sales funnel with organized, qualified opportunities.
What changed in the operation
- Inquiries are answered and quoted outside business hours.
- The advisor gets the inquiry with destination, date, passengers, and quote already loaded.
- Opportunities stop getting lost to slow replies.
- Departure and payment reminders are scheduled and go out on WhatsApp automatically.
Frequently asked questions about this case
Does the agent close the sale on its own?
No. It answers, quotes, and organizes the inquiry. The advisor closes the sale, stepping in when there is real intent and with the context already loaded.
Does it work on WhatsApp?
Yes, it is the agency's main channel. The agent answers on the same number and the team sees everything in the portal.
How does it avoid giving wrong information?
It answers from the catalog the agency maintains and hands off to a person when a question falls outside that scope.
Can the agent be switched off?
Yes. The portal has a control to turn the bot on or off, and each handoff is marked as handled once the advisor resolves it.
Keep exploring
Other work by the team, across different sectors.
Customer service · Team experienceAI chatbots for customer service in call centers
Voice and chat agents that resolve repetitive queries and hand off to the operator with context, without turning back into an IVR.- Intent understanding
- Handoff with history
- Voice and digital channels
Professional servicesAI assistants over WhatsApp for firms and professionals
One assistant per practice that takes the initial query, classifies the case, requests documents, and schedules before professional involvement.- One assistant per practice
- Configured by province
- Scheduling and reminders
FinanceArcos Capital: financial data traceability and automated reports
Financial reports generated automatically from Arcos Capital's sources, with every figure traced back to its origin.- Automated reports
- Every figure with its source
- Reconciliation and alerts
Tell us what you need to solve
In a 30-minute discovery call we review the process, the systems and the numbers, and tell you what's worth building, what isn't, and what return to expect. No cost, no commitment.