Menteia · Sommelier IA · October 2026 · Confidential
# Competition: customer complaints and features
**Sommelier IA** — AI voice sommelier and WhatsApp concierge that recommends pairings, quotes, books and follows up with every customer of restaurants, chefs, wineries and wine shops.
| Feature | Status | Priority | Phase | Note |
| --- | --- | --- | --- | --- |
| **Metrics dashboard: conversations, quotes, bookings and attributed sales** | Missing | High | Months 1–6 | Still pending even for the showcase client; needed to prove value against per-cover platforms. |
| **Deposits via payment link in WhatsApp** | Missing | High | Months 1–6 | Needed to close quoted events and reduce no-shows without the friction of forced prepayments. |
| **Age verification for alcohol sales** | Missing | High | Months 7–12 | Prerequisite for selling wine digitally; listed as a risk in the plan. |
| **Integration with reservation systems and POS** | Missing | Medium | Months 7–12 | Connect to OpenTable, SevenRooms, CoverManager or local POS (Fudo, Toteat) so the agent sees real availability and complements the tool the restaurant already uses. |
| **Table management and real-time availability** | Missing | Medium | Months 7–12 | Table-stakes in every reservation competitor; for restaurants without their own system. |
| **Wine catalog and sales over WhatsApp** | Missing | Medium | Months 7–12 | Planned in the roadmap (months 7–12) with a partner distributor; addresses the poor purchase experience of wine apps. |
| **English-language service for tourist destinations** | Missing | Medium | Months 19–24 | Planned for months 19–24; relevant in Valle de Guadalupe and beach destinations. |
| **Post-visit review requests and reputation** | Missing | Low | Months 13–18 | Common on reservation platforms; over WhatsApp it can be requested less intrusively than by email. |
| **Guest data owned by the business, exportable** | Partial | High | Months 1–6 | The CRM runs on its own database (Supabase), but it still needs to be documented, shown and offer full export (history and preferences). |
| **Multi-tenant template deployable in days** | Partial | High | Months 1–6 | It is the product promise, but today there is a single deployment; the roadmap turns it into a multi-tenant product in months 1–6. |
| **Human handoff and complex-request handling** | Partial | High | Months 1–6 | How the agent hands the conversation to the chef or manager is not documented; key to avoid the complaint about agents failing on unusual cases. |
| **Automatic WhatsApp confirmations and reminders** | Partial | High | Months 1–6 | The agent follows up, but there is no evidence of scheduled pre-booking reminders; competitors use them to cut no-shows. |
| **Repurchase campaigns segmented by preferences** | Partial | Medium | Months 13–18 | The CRM records preferences and occasions; campaigns are on the months 13–18 roadmap. |
| **24/7 WhatsApp concierge that quotes events and books** | We have it | — | — | Live in the showcase client; no reservation competitor offers it natively over WhatsApp with event quoting. |
| **Voice sommelier with pairing and brand tone** | We have it | — | — | Addresses complaints about robotic/generic voice and experiences without brand identity. |
| **Flat subscription with no per-cover fee** | We have it | — | — | US$150–300/month plans; the only commission is optional, on attributed wine sales. |
## To the roadmap
### Months 1–6
- **Guest data owned by the business, exportable** (partial, high): The CRM runs on its own database (Supabase), but it still needs to be documented, shown and offer full export (history and preferences).
- **Metrics dashboard: conversations, quotes, bookings and attributed sales** (missing, high): Still pending even for the showcase client; needed to prove value against per-cover platforms.
- **Multi-tenant template deployable in days** (partial, high): It is the product promise, but today there is a single deployment; the roadmap turns it into a multi-tenant product in months 1–6.
- **Human handoff and complex-request handling** (partial, high): How the agent hands the conversation to the chef or manager is not documented; key to avoid the complaint about agents failing on unusual cases.
- **Deposits via payment link in WhatsApp** (missing, high): Needed to close quoted events and reduce no-shows without the friction of forced prepayments.
- **Automatic WhatsApp confirmations and reminders** (partial, high): The agent follows up, but there is no evidence of scheduled pre-booking reminders; competitors use them to cut no-shows.
### Months 7–12
- **Age verification for alcohol sales** (missing, high): Prerequisite for selling wine digitally; listed as a risk in the plan.
- **Integration with reservation systems and POS** (missing, medium): Connect to OpenTable, SevenRooms, CoverManager or local POS (Fudo, Toteat) so the agent sees real availability and complements the tool the restaurant already uses.
- **Table management and real-time availability** (missing, medium): Table-stakes in every reservation competitor; for restaurants without their own system.
- **Wine catalog and sales over WhatsApp** (missing, medium): Planned in the roadmap (months 7–12) with a partner distributor; addresses the poor purchase experience of wine apps.
### Months 13–18
- **Repurchase campaigns segmented by preferences** (partial, medium): The CRM records preferences and occasions; campaigns are on the months 13–18 roadmap.
- **Post-visit review requests and reputation** (missing, low): Common on reservation platforms; over WhatsApp it can be requested less intrusively than by email.
### Months 19–24
- **English-language service for tourist destinations** (missing, medium): Planned for months 19–24; relevant in Valle de Guadalupe and beach destinations.
## Conclusion
The market complains about paying per cover, not owning its data and getting generic experiences or robotic voices; Sommelier IA addresses that with a flat subscription, native WhatsApp, brand-toned voice and wine expertise no reservation platform offers. Its biggest gap is operational: it still needs metrics proving results, deposit collection and integration with real table availability, and it currently relies on a single showcase client.