Problem
Newcrux solves e-commerce friction in Latin America — the friction of forms, passwords, registrations, and dozens of clicks that causes 87% of the carts are abandoned. That same friction rules out millions of people who are unable or unwilling to navigate complex interfaces, especially the more than 24 Millions of Brazilians 60+ connected people who face digital barriers. The result is twofold: retailers lose sales every day, and a huge portion of the population is dependent on third parties for a task as simple as buying. Newcrux eliminates this barrier by transforming the purchase into a natural voice or text conversation on WhatsApp, the web, or the app.
Solution
Newcrux exchanges the traditional e-commerce interface - forms, passwords and dozens of clicks - for a natural conversation, by voice or text, on WhatsApp, on the web or on the app. The core is the orchestration of three IAs: the semantic understands what the customer wants, even in colloquial language; the generative one answers and drives the conversation; and the agency closes the sale — consults the stock, sets up the order and integrates the payment into the same conversation. All natively connected to the retailer's catalog and catwalk, without installing anything or creating an account. The customer says “I want two kilos of rice” by audio and buys it right there. Thus, Newcrux recovers lost sales in 87% of abandoned strollers and restores autonomy to those who previously relied on a family member or gave up digital.
Business model
Phase 1 — Today: Project + Implementation. One-time fee for integration and initial agent configuration. Ticket varies according to the complexity and size of the customer. It's what funds the setup and validates each retailer. Phase 2 — 6 months: Monthly subscription (SaaS). Recurring plan by volume of conversations or by active store. Market Benchmark: US$ 39–139/month. MRR Goal: US$ 50K. Phase 3 — 18 months: Platform + Revenue Share. Self-serve access to the platform plus a commission on sales closed by the AI agent. It's the model that scales without friction. ARR goal: US$ 4,7M. For the average ticket, the number that supports the deck's SOM is ~US$ 300/month per store (1% from 130 thousand stores × US$ 300/month = US$ 4,7M for ARR). The cited competitive benchmark: Tidio US$ 39/month, ManyChat US$ 15–169/month, Botmaker enterprise — Newcrux is positioned in the mid-market LATAM with deep integration.
Market
TEAM — US$ 769 billions, the total e-commerce in Latin America in 2025, growing 21% per year and designed to surpass US$ 1 Trillion in 2027 (source: PCMI) 2025). SAM — US$ 33 billions, e-commerce in Argentina, the market with the highest CAGR in the region (14% unto 2027), as an entry point. SOUND — ~US$ 4,7M from ARR, capturing 1% Das 130 thousand active online stores × ~US$ 300/month.
Competitors
1. Chatbots and traditional service platforms (Zenvia, Take Blip, ManyChat, Zendesk). They answer questions through strict decision trees, but they don't close the sale with inventory and payment in the same conversation. 2. E-commerce and checkout platforms (VTEX, Nuvemshop, Tray, Shopify). They optimize the traditional funnel with fewer clicks, but they maintain the logic of the form and interface - they don't replace the click with the conversation. 3. Human service via WhatsApp. The strongest competitor: it solves for those who have digital difficulties, but it costs USD 8–15 for service and it doesn't scale. This is the space that Newcrux occupies - the human experience of WhatsApp with the cost and scale of an agency AI (USD). 0,50–0,70).
Competitive differentiation
The difference between Newcrux is not that it is “another conversational AI” - it is that it is the AI infrastructure for the complete sales cycle in food retail. While the chatbot only responds and the platform only processes the checkout, Newcrux orchestrates the entire journey in the conversation: it understands the order, consults the stock, sets up the cart, closes the payment and organizes the repurchase. The second differential, and the most defensible, is the data. Each conversation captures intent, preference, and buying behavior in natural language—not just what the person clicked, but what they ordered and what they stopped buying. This becomes inventory predictability, customer reactivation, and a proprietary data asset that no chatbot or checkout platform can build, because they are not present at the time of the decision. In short: competitors compete for parts of the funnel; Newcrux unifies conversational selling and data intelligence in a single layer, designed for the supermarket operation.
Entry barrier
The main barrier is the proprietary data that accumulates with use. Each conversation feeds a base of intention and behavior when buying food in LATAM that is neither bought nor copied — it is only built with time and volume. The more Newcrux operates, the more accurate it is in its inventory forecast, recommendation and reactivation, increasing the distance for those who arrive later. The second barrier is deep integrations in the retailer's operation: inventory, catalog, and payment connected natively. Once the supermarket runs its sales cycle on this layer, the exchange cost is high - it does not replace the infrastructure that already sells without risking the operation. Both are reinforced: data improves the product, the product deepens integration, and integration generates more data. A new entrant would need to replicate all three things at the same time, starting from scratch in conversation volume.
Traction
Yes. The traction that Newcrux generated was real and verifiable in production - not projection. From the pitch: US$ 110K of revenue in the last 12 months, US$ 18.500 of stable MRR (Feb/2026), with 1 large customer with active mass consumption, with branches across the country. Marcos: foundation in mar/2025 with US$ 20K of the founders; MVP in Jun/2025; onboarding the large customer in ago/2025; stable monthly revenue of US$ 18,5K starting at ten/2025; Google Chile reference in Feb/2026 and selection for the Emprelatam accelerator. The point of honesty, and the update for today's date: that traction has been discontinued. It was real revenue, in production, with a large customer paying on a recurring basis — but the contract/operation was closed, and the MRR of US$ 18,5K is no longer active. In other words: Newcrux has already proven that it can sell, integrate, and sustain recurring revenue with a large customer; what was cut was the continuity of that specific account, not the ability to generate it.