<div class="portada"><div class="meta">Menteia · RosettaDMS · October 2026 · Confidential</div>

# Business plan

**RosettaDMS** — Connects a car brand with all of its dealers whatever their DMS: through an API or an AI agent that operates it like a service advisor, a central brain that translates everything, an MCP server and voice with Allo.</div>


## 1. Summary
RosettaDMS connects each dealer with the best available method: if its DMS has an API it uses it; if not, a computer-use AI agent logs in with a dedicated user and operates the system like a service advisor (reads the screen, navigates menus, downloads reports and books appointments). Everything flows into a central brain that translates it into a single schema (Rosetta Schema) with history by VIN, and from there out to the brand (console, data lake, alerts), to AI agents through an MCP server, and to a voice receptionist with Allo that answers and books against the shop's real schedule. The dealer pays nothing and installs nothing.

> All your dealers. Any DMS. One language.

## 2. Problem
- Each dealer group picked its DMS years ago: a brand with hundreds of points of sale lives with dozens of different systems and cannot see service, parts or inventory in real time.
- Most DMS in Mexico have no open API; integrating each one takes months and depends on its vendor.
- The end customer pays for the fragmentation: nobody answers to book, the part is not there on arrival and the car's status is not real.

## 3. Solution and product
- Connect: API connector (REST/SOAP, SFTP, webhooks) or computer-use connector on an isolated virtual desktop that records every session; hybrid mode when the API does not cover everything.
- Translate: Rosetta Schema unifies 'OT', 'RO', 'Orden' and 'Work order' into the same entity; full history by VIN even when the car visits several dealers.
- Think: Just-in-Time parts demand, early quality alerts by model, batch or region, and dealers at risk.
- Act: writes appointments, leads and parts reservations into the DMS with prior validation; automatic rules (lead SLA, 'car ready' notice).
- Expose: own MCP server (buscar_disponibilidad, agendar_cita, estatus_orden, existencia_parte, inventario_unidades, historial_vin) and a voice and SMS layer with Allo.

### Technology
- Computer-use agents with a playbook per DMS and vision-based adaptation when a screen changes.
- Reusable API adapters and a common data schema with history by VIN.
- Own MCP server; integration with Allo (API, webhooks and its own MCP) for voice, SMS and call summaries.
- Encrypted credential vault, recorded sessions, pseudonymized data toward the brand (Mexican data protection law).

## 4. Market
| Level | Annual size | How it is calculated |
| --- | --- | --- |
| TAM | US$10 M | ~2,800 car dealerships in Mexico (AMDA) × ~US$3.6k/year per connected dealer |
| SAM | US$4 M | Brands with 100+ dealers and mixed DMS (~1,100 dealers of the highest-volume brands) |
| SOM | US$500k | ~150 connected dealers from 1–2 brands in 24–36 months at ~US$3.6k/year |

**Supporting data:**
- More than 2,800 dealerships of 75 brands in 210 cities (AMDA).
- 1.52 M new cars sold in 2025 (INEGI); 1.62 M according to AMDA.
- The site cites the typical case of a brand with 200+ points of sale and 9 different DMS (an illustrative example, not a customer).

*Market figures are estimates based on public sources; validate them before sending to institutional funds.*

## 5. Target customer
- Car makers and brands (OEMs) with dealer networks in Mexico.
- Dealer groups with several DMS across their stores.
- Later: multi-country brands in LatAm, motorcycles, trucks and machinery.

## 6. Business model and pricing
SaaS paid by the brand per connected dealer; the dealer never pays. 30-day pilot with 4 dealers on 4 different DMS. Allo's voice license is contracted separately and the brand can sponsor it. Per-dealer prices are not public yet: the ones in this table are a proposal for the model.

| Item | Price |
| --- | --- |
| Pilot (30 days, 4 dealers, 4 DMS) | Free for the dealer; price to the brand to be defined |
| National Network: unlimited dealers, hourly sync | Proposed: ~US$300 per dealer per month |
| Enterprise: deployed in the brand's cloud, SSO, SLA | Custom |
| Voice with Allo | Separate license (independent provider) |

## 7. Go-to-market strategy
- Pilot program: 4 dealers on different DMS, 30 days, free for the dealer.
- Sell to the brand (OEM), not dealer by dealer: the brand funds the network and recommends it to its dealers.
- Entry through Menteia's automotive network (Vamo is piloting with Ford/Lincoln dealerships).
- Every mapped DMS becomes reusable: the next dealer on that DMS connects in minutes.

## 8. Competition and positioning
| Alternative | What it offers | Our advantage |
| --- | --- | --- |
| DMS vendors (CDK, Reynolds, Quorum, local DMS) | Integration only with their own system | Unifies every DMS in the network into one schema |
| Automotive data integrators (e.g. Authenticom/DealerVault) | DMS data extraction, US-focused | Reads and writes, works without an API and exposes MCP for agents |
| Generic RPA (UiPath, Automation Anywhere) | Brittle bots, one project per customer | Reusable per-DMS playbooks and vision-based adaptation |
| Excel reports and brand portals | Late data and double entry | Daily or hourly data with no manual entry |

## 9. Defensible advantages
- Works with DMS that have no API (desktop, AS-400 terminals, Citrix/RDP, in-house systems), which are the majority in Mexico.
- No cost and no installation for the dealer: the brand pays per connected dealer.
- Bidirectional: it does not just read reports, it also books and reserves parts inside the DMS.
- Built for AI agents: any bot, call center or voice receptionist can act across the whole network via MCP.

## 10. Traction
- Site and interactive demo live at rosettadms.menteia.com (console with fictional dealers and figures).
- Pilot program defined: 4 dealers on 4 different DMS in 30 days.
- Allo integration designed (MCP, webhooks and API); no customers or running pilot yet.

## 11. Team
| Partner | Role | Stake in Menteia |
| --- | --- | --- |
| Rafael Herrera | CEO / CFO | 33.3% |
| Vladimir Díaz | CTO | 33.3% |
| Francisco D'Angelo | Head of Legal and Compliance | 33.3% |

**Hires funded by the round:**
- Senior full-stack engineer — from month 1
- B2B account executive — from month 5
- Customer success / implementation — from month 12

## 12. Financial projection (24 months)
| Quarter | Revenue | Expenses | Net cash flow | Cumulative cash flow |
| --- | --- | --- | --- | --- |
| Q1 | US$0 | US$34,965 | US$-34,965 | US$-34,965 |
| Q2 | US$2,400 | US$41,127 | US$-38,727 | US$-73,691 |
| Q3 | US$11,700 | US$44,370 | US$-32,670 | US$-106,361 |
| Q4 | US$36,000 | US$47,248 | US$-11,248 | US$-117,608 |
| Q5 | US$60,300 | US$52,680 | US$7,620 | US$-109,988 |
| Q6 | US$84,600 | US$53,004 | US$31,596 | US$-78,392 |
| Q7 | US$108,900 | US$53,328 | US$55,572 | US$-22,819 |
| Q8 | US$131,100 | US$53,652 | US$77,448 | US$54,629 |

- Monthly recurring revenue (MRR) at month 24: **US$45,000** (ARR US$540,000)
- Total 24-month spend: **US$380,371**; cumulative revenue: **US$435,000**
- Projected monthly breakeven: **month 14**

**Revenue assumptions:**
- No revenue during the pilot until month 4; the 4 pilot dealers pay from month 5.
- From month 8, ~9 dealers per month from the brand's network: ~150 by month 24.
- Proposed price: US$300 per connected dealer per month, paid by the brand.
- AI includes compute for computer-use agents and virtual desktops.

## 13. The round
- Amount: **US$425,000** for **29.8%** (pre-money US$1,000,000, post-money US$1,425,000).
- Use of funds detailed in *06_Modelo_Financiero_y_Ronda* (financial model and round).

## 14. Risks and mitigation
| Risk | Mitigation |
| --- | --- |
| No pilot or customers yet | 30-day pilot free for the dealer; first brand through Vamo's network |
| DMS licenses that forbid automated users | Legal review per DMS; prefer the API when there is one; the dealer creates and revokes the user |
| Computer use is brittle when screens change | Versioned playbooks, vision-based adaptation and monitoring of every sync |
| Long sales cycle with car makers | Start with dealer groups and a mid-size brand |
| Dependence on Allo for voice | Voice is a module; the core works without it and there are alternative providers |