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

# Business plan

**Copiloto Aero** — AI quoting cockpit for aerospace suppliers: reads requests and drawings, prepares the quote with the company's pricing engine, and the analyst approves every section.</div>


## 1. Summary
Copiloto Aero turns every request email into a quoting session: it triages the drawing with computer vision, looks up past quotes, runs the company's pricing engine and presents each section with its rationale so the analyst can approve, ask or adjust by text or voice. It generates the PDF and the email to the customer, but nothing goes out without human approval and the LLM never makes up a number.

> Quote in hours what takes days today, without losing control.

## 2. Problem
- Aerospace suppliers receive requests for quotation (RFQs) with drawings, specifications and materials; preparing each one takes an expert analyst days.
- The knowledge needed to quote lives in the heads of a few people and in thousands of past quotes that nobody looks up.
- Part number, material or tolerance errors cost orders or margin; and generic AI is not reliable with numbers.

## 3. Solution and product
- Each RFQ email becomes a session linked to the CRM (Pipedrive) and the company's quoting engine.
- Drawing triage with vision: detects discrepancies (e.g., the part number in the email differs from the one on the drawing).
- Knowledge base of past quotes (929 quotes, 4,600 line items at the anchor customer).
- The analyst approves, asks or adjusts each section by text or voice; human gates at technical review, pricing and sending.
- Customer-branded quote PDF, clarification email to the customer and a global command dashboard.

### Technology
- Python 3.12 + FastAPI (~5.8k lines), OpenAI-compatible endpoint, 18 test files.
- Open WebUI as a white-label cockpit; agent with 26 tools on the OpenAI API (text, vision, real-time voice).
- Supabase (Postgres + pgvector); Pipedrive, read-only IMAP and SMTP integrations.
- Docker + nginx on a VPS.

## 4. Market
| Level | Annual size | How it is calculated |
| --- | --- | --- |
| TAM | US$45 M | ~3k precision manufacturing and distribution suppliers in Mexico that quote against drawings (aerospace, medical, premium automotive) × US$15k/year |
| SAM | US$5.8 M | 386 aerospace companies in Mexico × US$15k/year |
| SOM | US$250k | ~11 companies in 24 months with a ticket of ~US$22k/year |

**Supporting data:**
- Mexico's aerospace exports 2025: ~US$13.6 B (record); 4th largest exporter to the US.
- 386 aerospace companies in 19 states (1H-2024); FEMIA groups ~130 (85% of the sector).
- Domestic market of US$11.2 B in 2025 with a target of US$22.7 B in 2029 (FEMIA).
- MRO: the fastest-growing projected segment (8.35% per year).

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

## 5. Target customer
- Tier 2 and Tier 3 aerospace suppliers (machining, assemblies, parts) in Querétaro, Chihuahua, Sonora, Baja California and Nuevo León.
- Aerospace parts and materials distributors and MRO shops.
- In the future: medical and automotive precision manufacturing that quotes against drawings.

## 6. Business model and pricing
One-time implementation (integration with email, CRM and pricing engine + ingestion of historical data) and a monthly subscription per company based on users and quote volume.

| Item | Price |
| --- | --- |
| Implementation and historical data ingestion | US$8,000–15,000 one-time |
| Subscription per company (up to 10 users) | US$1,500 / month |
| Additional users or volume | US$100 / user / month |
| Pricing engine module (if the company has none) | Quoted separately |

## 7. Go-to-market strategy
- Anchor case: INDUMET (aerospace quoting) in production as a reference.
- Direct sales in aerospace clusters (Querétaro, Chihuahua, Sonora, Baja California) and through FEMIA.
- Industry trade shows (FAMEX, Aerospace Meeting) and partnerships with industrial ERP/CRM integrators.
- Expansion to US suppliers through Mexican customers (nearshoring).

## 8. Competition and positioning
| Alternative | What it offers | Our advantage |
| --- | --- | --- |
| Manual quoting (Excel + email) | What they do today | Days → hours, with history and traceability |
| Industrial ERP / CPQ (SAP, Paperless Parts) | Expensive, rigid configurators | Sits on top of the existing CRM and email |
| Generic AI chat | Answers, but makes up numbers | Zero LLM numbers; human at every gate |
| Custom consulting | Long projects | Ready product in weeks |

## 9. Defensible advantages
- Zero LLM-generated numbers: every figure comes from the pricing engine with its source; the AI explains.
- Human at every critical gate (technical, pricing, sending): designed for regulated industries.
- Learns from the company's history: past quotes as context, with opt-in learning.
- Sits on top of what the customer already uses (email, CRM, pricing engine) instead of replacing it.

## 10. Traction
- In production at INDUMET (aerospace quoting): real inbox, live CRM and historical data ingested.
- 929 past quotes, 4,600 line items and 29 historical customers as the knowledge base.
- 7 INDUMET team members operating the cockpit.

## 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:**
- AI / ML engineer — from month 1
- Senior full-stack engineer — from month 3
- B2B account executive — from month 4
- Customer success / implementation — from month 7

## 12. Financial projection (24 months)
| Quarter | Revenue | Expenses | Net cash flow | Cumulative cash flow |
| --- | --- | --- | --- | --- |
| Q1 | US$1,700 | US$39,551 | US$-37,851 | US$-37,851 |
| Q2 | US$16,800 | US$61,624 | US$-44,824 | US$-82,675 |
| Q3 | US$23,600 | US$69,556 | US$-45,956 | US$-128,632 |
| Q4 | US$42,100 | US$69,826 | US$-27,726 | US$-156,358 |
| Q5 | US$38,900 | US$70,096 | US$-31,196 | US$-187,554 |
| Q6 | US$57,400 | US$70,366 | US$-12,966 | US$-200,520 |
| Q7 | US$54,200 | US$70,636 | US$-16,436 | US$-216,956 |
| Q8 | US$72,700 | US$70,906 | US$1,794 | US$-215,163 |

- Monthly recurring revenue (MRR) at month 24: **US$28,700** (ARR US$344,400)
- Total 24-month spend: **US$522,563**; cumulative revenue: **US$307,400**
- Projected monthly breakeven: **month 18**

**Revenue assumptions:**
- INDUMET moves to a subscription in month 3 (US$1,700/month).
- One new customer every 2 months from month 6: ~11 customers by month 24.
- US$10,000 implementation per new customer, billed in its sign-up month.
- Average ticket of US$1,700/month (subscription + additional users).

## 13. The round
- Amount: **US$575,000** for **24.2%** (pre-money US$1,800,000, post-money US$2,375,000).
- Use of funds detailed in *06_Modelo_Financiero_y_Ronda* (financial model and round).

## 14. Risks and mitigation
| Risk | Mitigation |
| --- | --- |
| Single customer and dependence on its (third-party) pricing engine | Generic pricing engine connector and our own costing module |
| Instance domain (.org vs .com) | Confirm with the customer; own brand for new customers |
| Accidental sending to the end customer | Human gates and testing of stage rules with the engine provider |
| Dependence on Open WebUI (license to remove its branding) | Enterprise license or our own cockpit |
| Security of technical data (ITAR/export controls) | Dedicated server per customer, encryption and access controls |