TallerYa

1. One-liner

WhatsApp AI copilot that runs quotes, invoices, and CFDI compliance for Mexico’s 260K independent garages.

2. Trend signal — why now?

Three forces are converging on Mexico’s independent mechanics right now:

Regulatory squeeze. Mexico’s SAT published new CFDI reform rules for 2026 that explicitly target false or simulated invoices. Starting May 2026, technology platforms must give SAT real-time access to transaction data. Shops that were casual about invoicing now face real consequences — CSD suspension, sanctions, loss of tax effects. Every taller mecánico in the country needs to get their CFDI house in order, and most don’t have the tools to do it.

WhatsApp-as-business-OS. Mexico has 74M WhatsApp users at 93% penetration. 78% of Mexican WhatsApp users already interact with businesses through the app. These mechanics are already sending photos and quotes to customers over WhatsApp — they just can’t generate invoices, track jobs, or manage inventory from the same chat thread.

AI cost collapse. GPT-4o-level inference is now under $1 per 1M input tokens. A WhatsApp-native AI copilot that understands Spanish automotive terminology, generates quotes from voice notes, and auto-fills CFDI fields is now buildable for pocket change. A year ago, the inference cost alone would have eaten the margin.

Provenance:

• Signal 1: SAT 2026 CFDI reform tightens invoicing requirements for all Mexican businesses, including auto repair shops — https://blog.gigstack.pro/post/reforma-fiscal-2026-nuevas-reglas-cfdi-sat — 2025-12-30
• Signal 2: DENUE 2025 registers 261,806 auto repair economic units in Mexico, majority independent — https://www.economia.gob.mx/datamexico/en/profile/industry/automotive-repair-and-maintenance — 2025
• Signal 3: Mexico has 93% WhatsApp penetration (74M users), 78% interact with businesses on the platform — https://mazkara.studio/en/newsletter/whatsapp-penetration-latin-america-2026/ — 2026 Category: Geographic arbitrage + Regulatory arbitrage

3. The opportunity

Mexico has 261,806 registered auto repair shops (DENUE 2025). Across LATAM, there are 839,000 shops, 81% of which are independent. These aren’t franchise dealerships — they’re 1–3 person operations with a WhatsApp number on a cardboard sign.

The incumbents are misaligned:

• Mecanica MX ($799 MXN/mo) is a web-first dashboard with WhatsApp notifications bolted on. It assumes the mechanic sits at a computer, which they don’t — they’re under a car.
• AutoSoft Taller (~$31 USD/mo) is a Windows desktop application. In 2026. For a guy whose only computer is his phone.
• US tools (Shopmonkey, Tekmetric, AutoLeap) are English-only, $200-400/mo, and designed for 5+ bay shops with service advisors. Completely wrong fit.

The gap is a WhatsApp-native tool where the mechanic never opens a browser or downloads an app. They text their copilot a photo of the problem, get a quote template, send it to the customer for approval in the same chat, and generate a CFDI invoice when the job’s done — all from the phone that’s already in their grease-stained hand.

4. Target market

• Primary customer: Independent mechanics and small garage owners in Mexico, typically 1–3 employees, monthly revenue MX$50K–200K (~$2,800–$11,000 USD), concentrated in Estado de México (32K shops), Jalisco (17K), and Veracruz (16K).
• Why they buy: They need to issue CFDI invoices to stay compliant with SAT, but existing software requires a computer they don’t have or costs more than they can justify. They’re already sending quotes over WhatsApp — they just want the quote to become an invoice without retyping everything.
• Rough TAM reasoning: 261,806 shops in Mexico × even 1% penetration × $199 MXN/mo × 12 = ~$6.2M MXN ARR ($345K USD). At 5% penetration: $1.7M USD ARR. Expand to Colombia, Argentina, and the rest of LATAM and the math comfortably clears $5M ARR.
• Why now for them: SAT 2026 reform. If your invoices can’t be proven as real transactions, you face CSD suspension. For shops that were handing customers handwritten receipts, the clock is ticking.

5. Product sketch (MVP)

• WhatsApp-native interface — mechanic texts or sends voice notes to the TallerYa number. No app download, no web dashboard required.
• AI quote generator — send a photo of the problem + a voice note describing the repair; TallerYa generates a structured quote with parts, labor estimate, and total in MXN.
• Customer quote approval — forward the formatted quote to the customer’s WhatsApp for tap-to-approve.
• CFDI invoice generation — on job completion, one message stamps a CFDI 4.0 invoice through a PAC (Facturama/FacturAPI) and sends the XML+PDF to the customer.
• Job tracking — simple status tracking (received → in-progress → ready → delivered → invoiced) managed through WhatsApp commands or quick replies.
• Parts cost log — mechanic logs parts purchased with photos of receipts; AI extracts amounts for cost tracking.
• Monthly summary report — automated WhatsApp message with total revenue, jobs completed, parts costs, and margin estimate.

6. AI angle — what’s load-bearing

AI does three irreplaceable jobs here:

1. Voice-to-quote. The mechanic sends a 15-second voice note (“Cambio de balatas delanteras, disco derecho rayado, rotación de llantas”) and AI parses it into a structured quote with line items, pulling from a labor-rate reference database and local parts pricing. Without AI, the mechanic would need to type all of this on a phone keyboard with greasy fingers — which is why they currently just send an unformatted “te sale en 3500” text.

2. Receipt OCR → cost tracking. Photos of handwritten or printed parts receipts get parsed into structured cost entries. No mechanic is going to manually enter receipt data. Without AI, there’s no cost tracking at all.

3. CFDI auto-fill. AI maps quote line items to SAT product/service codes (clave de producto/servicio), auto-fills RFC fields, and generates the CFDI XML. This is the compliance piece that’s currently done by an accountant charging $200-500/mo or by the mechanic painfully navigating the SAT portal.

Remove the AI and you’re back to “type your invoice on a website,” which is what the incumbents already offer and which these mechanics already don’t use.

7. Localization angle

This is a Mexico-first play by design.

• Language: 100% Spanish. Automotive slang is highly regional — “balatas” (Mexico) vs “pastillas” (Spain/Colombia). AI must speak like a Mexican mechanic, not a translation model.
• Payment rails: MXN pricing, SPEI/CoDi/cash. No credit card required for the mechanic to pay for TallerYa — OXXO or SPEI transfer works.
• Compliance: CFDI 4.0 with PAC integration is Mexico-specific. This is the moat — a US tool can’t touch this.
• Pricing: MX$199–499/mo ($11–28 USD). Existing alternatives charge $800 MXN/mo and above. A sub-$300/mo WhatsApp tool that handles invoicing will feel like a no-brainer.
• Distribution: WhatsApp is the medium. You reach mechanics where they already are — no app store, no Google Ads needed.
• Expansion: Colombia (electronic invoicing mandatory since 2019), Argentina, Peru all have similar profiles — independent garages, WhatsApp dominance, local e-invoicing mandates. Each requires a new compliance module but the core WhatsApp+AI engine is reusable.

8. Business model — path to $1M–$5M ARR

• Pricing: MX$299/mo base plan ($17 USD), MX$499/mo pro with unlimited CFDI and parts tracking ($28 USD). Per-invoice fee of MX$3 (~$0.17) for pay-as-you-go users.
• ACV: ~MX$3,600–6,000/year per shop ($200–335 USD)
• Rough math to $1M ARR: 3,500 shops × $285 ACV = ~$1M USD. That’s 1.3% of Mexico’s 261K shops.
• Rough math to $5M ARR: 10,000 shops in Mexico + 7,500 across Colombia/Argentina/Peru = 17,500 × $285 = $5M USD.
• Expansion path: Upsell parts ordering marketplace (affiliate revenue from auto parts distributors), insurance claim integration, and fleet management for multi-bay shops at higher tier pricing.

9. Go-to-market wedge — first 100 customers

1. Refaccionarias (parts stores) as channel partners. Auto parts stores are the mechanic’s daily touchpoint. Partner with 10 refaccionarias in Estado de México — the state with 32K shops — and offer a printed flyer with a WhatsApp QR code: “Envía ‘HOLA’ y prueba gratis 30 días.” Each refaccionaria serves 50–100 mechanics. 10 stores × 5% conversion = 25–50 shops in month one.

2. DENUE scrape + targeted WhatsApp outreach. DENUE provides name, address, and phone for all 261K registered shops. Scrape the top 5,000 in Estado de México and Jalisco, send a personalized WhatsApp message: “¿Emites facturas CFDI? TallerYa te ayuda desde WhatsApp por $299/mes. Prueba gratis.” At 3% conversion: 150 leads, 50 trials, 15 paying.

3. YouTube/TikTok content in Spanish. Short-form videos showing a real mechanic using TallerYa to go from voice note to CFDI invoice in 60 seconds. Target “facturación taller mecánico” and “cómo facturar en mi taller” keywords — low competition, high intent. Goal: 5–10 organic inbound leads per week by month 3.

4. CNT partnership. The Confederación Nacional de Talleres de Servicio Automotriz is the industry association. A partnership or endorsement gives instant credibility and access to their member base.

5. Referral engine. Every CFDI invoice sent through TallerYa includes a “Powered by TallerYa” footer with a link. Customers who receive professional invoices ask their other mechanics why they can’t do the same. Organic viral loop built into the invoicing flow.

10. Build complexity — justification

Medium. The WhatsApp Business API integration is well-documented. CFDI PAC integration (Facturama or FacturAPI) has REST APIs with SDN libraries. LLM inference for Spanish voice-to-text and quote parsing uses off-the-shelf models (Whisper + GPT-4o-mini). The hard parts are: (a) building a robust SAT product code mapping for automotive services, and (b) handling edge cases in voice note parsing for Mexican automotive slang. Realistic timeline: 10–14 weeks for a 2-person team to reach MVP with basic quoting + CFDI invoicing.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Invoicing assistance is legal. PAC handles CFDI certification.
Ethical — no harm / dark patterns	✅	Helps formalize informal businesses. Net positive for tax compliance.
Market exists (evidence above)	✅	261K registered shops. Incumbents charging money. SAT deadline forcing action.
1–5 person team can build this	✅	2–3 engineers. WhatsApp API + LLM + PAC API. Standard stack.
Launchable with <$50K / ₹40L	✅	Estimated launch cost: $15–25K USD. WhatsApp API costs negligible at low volume.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	SAT compliance is a real threat. Mechanics lose customers who need invoices. But many shops have survived informally for years — urgency varies.
Demand evidence	15	12/15	261K shops, multiple incumbents charging money, SAT reform creating urgency. No direct “I wish there was a WhatsApp invoicing tool” signal, but the behavior is there — they already quote over WhatsApp.
Build feasibility	15	12/15	Standard APIs throughout. Voice-to-quote parsing needs tuning for automotive Spanish. CFDI mapping is grunt work but not technically hard. 10–14 weeks.
Distribution clarity	15	12/15	Refaccionarias + DENUE scrape + WhatsApp outreach is concrete. CNT partnership is aspirational but realistic.
Revenue mechanics	15	11/15	$17–28/mo is affordable for the target. But collecting payments from informal businesses in Mexico requires OXXO/SPEI support — no credit card dependency. Churn risk if SAT enforcement is uneven.
Time to first revenue	10	7/10	30-day free trial means revenue in ~6–8 weeks post-launch. Conversion depends on trial-to-paid in a price-sensitive market.
Defensibility	10	6/10	CFDI integration + automotive SAT code mapping creates a compliance moat. WhatsApp conversation history becomes switching cost. But any well-funded competitor could replicate in 6 months.
Total	100	76/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy — needs strong backend engineering for WhatsApp API + PAC + LLM integration. Spanish-speaking technical founder is ideal. Sales happens through the product and channel partners, not enterprise sales calls.

Key assumptions to validate (3–5)

1. Assumption: Independent mechanics will pay MX$299/mo for a WhatsApp invoicing tool. How to test: 50 cold WhatsApp messages to DENUE-listed shops in Estado de México offering a free 30-day trial. Measure trial signup rate and day-30 conversion.
2. Assumption: Voice-to-quote accuracy is good enough for Mexican automotive terminology. How to test: Record 50 real voice notes from mechanics describing common repairs. Measure AI parsing accuracy against manually created quotes.
3. Assumption: Refaccionarias will distribute flyers and QR codes for a revenue share. How to test: Visit 10 refaccionarias in Ecatepec/Toluca and pitch the partnership. Measure willingness and terms.
4. Assumption: SAT enforcement of the 2026 reform will be strong enough to drive adoption. How to test: Monitor SAT announcements and interview 20 mechanics about their current compliance status and concern level.

Risk flags

1. [Regulatory enforcement risk]: If SAT enforcement of the 2026 CFDI reform is weak or delayed, the urgency driver evaporates. Mechanics who’ve been informal for 20 years won’t change habits without real consequences.
2. [Platform dependency]: 100% reliance on WhatsApp Business API. Meta can change pricing, terms, or rate limits. Mitigate by building the core logic platform-agnostic with WhatsApp as a channel adapter.
3. [Payment collection risk]: Collecting monthly SaaS fees from informal Mexican businesses without credit cards is operationally complex. OXXO cash payment integration (via Conekta/Stripe Mexico) is essential but adds friction.

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Spanish-speaking technical founder with Mexico market access
Time to revenue:        8–10 weeks post-launch
Capital to launch:      $15–25K USD (~MX$270K–450K)
Top 3 assumptions to validate first:
  1. Mechanics will pay $299 MXN/mo — test with 50 cold WhatsApp outreaches + free trial
  2. Voice-to-quote accuracy for Mexican automotive Spanish — test with 50 recorded voice notes
  3. Refaccionarias will act as distribution channel — test with 10 in-person visits
Kill criteria:
  - Abandon if <5% of 50 trial signups convert to paid after 30 days
  - Abandon if voice-to-quote accuracy <70% after 2 weeks of fine-tuning
  - Abandon if 0 of 10 refaccionarias agree to distribute flyers

15. Next step — 1-week validation sprint

• Day 1: Scrape 200 shops from DENUE in Estado de México with phone numbers. Set up a WhatsApp Business account. Prepare a 30-second Spanish-language pitch video showing the concept (mockup, not working product).
• Day 2: Send personalized WhatsApp messages to 100 shops: “¿Ocupas facturar en tu taller? Estamos construyendo una herramienta que te permite cotizar y facturar desde WhatsApp. ¿Te interesaría probarla gratis?” Track reply rate and sentiment.
• Day 3–4: Visit 5 refaccionarias in Ecatepec and Toluca. Ask: (a) would you put a flyer on your counter? (b) how many mechanics come in daily? (c) do your mechanics ask about invoicing tools? Record answers.
• Day 5: Analyze: if >15% of WhatsApp messages get a positive reply AND 3+ refaccionarias say yes to distribution, proceed to build. If either fails, pivot or PASS.

The validation is binary: either mechanics respond to the WhatsApp pitch and parts stores will distribute, or they don’t. No “people seemed interested” — count the replies.