Robbin integrates credit intelligence with Claude and ChatGPT via MCP
Robbin, a São Paulo-based credit intelligence company, has launched a Model Context Protocol integration that lets finance teams query its credit-scoring engine directly inside Claude and ChatGPT. The company says it is the first MCP-based credit product in Latin America, giving credit analysts and CFOs the ability to request limit recommendations, risk assessments and buyer portfolio reviews through a conversational interface rather than switching between spreadsheets and separate reporting tools.
The launch follows Robbin's $108 million financing announced four months earlier: an $8 million seed round co-led by Canary, Atlântico and Caravela, and a $100 million receivables investment fund (FIDC) structured with Augme, the asset management arm of XP Investimentos. The seed capital is funding platform development and AI product work; the FIDC is ring-fenced for credit operations.
What the product does
Robbin's MCP integration connects to the company's proprietary credit model, which it calls the Robbin score. When a credit analyst asks for a limit recommendation on a specific reseller, the AI assistant pulls the buyer's purchase history with the manufacturer alongside broader market-behaviour data, then returns a suggested limit and risk flag within seconds. The company says each manufacturer's data remains isolated and is used solely to analyse that company's own customer base, which addresses a common objection to AI-assisted credit tools: cross-contamination of proprietary commercial data.
MCP is an open standard originally developed by Anthropic to allow AI assistants to reach external systems and databases in a controlled, auditable way. Its adoption by third-party vendors is accelerating: embedding MCP connectors into vertical software platforms is an emerging pattern across enterprise software, particularly for workflows where structured data retrieval is more valuable than free-text generation.
Leonardo Moura, chief executive of Robbin, said the shift AI delivers is "supporting the decisions that move the business," rather than automating routine tasks. He described the product as giving finance teams "a read on its base that it couldn't get alone."
Market context and competitive positioning
The addressable market for Robbin sits at the intersection of B2B trade credit, embedded finance and AI workflow tooling, a combination that remains under-served in Latin America despite the region's large informal distribution networks. In markets such as Brazil, manufacturer-reseller credit has historically been relationship-driven rather than data-driven, meaning incumbents tend to be internal credit teams and legacy ERP credit modules rather than specialist AI vendors.
Globally, the AI-powered credit decisioning space includes companies such as Credolab, Zest AI and various credit-bureau API providers, though none of these has announced MCP-native integrations with general-purpose AI assistants to date. If the open-MCP approach gains traction, it lowers the integration barrier significantly: a finance team already using Claude or ChatGPT Enterprise can add credit intelligence without a custom API project.
Regulatory considerations differ by market. Brazil's open finance framework, operationalised through the Banco Central do Brasil's open data directives, creates a structured environment for sharing financial data, but also imposes consent and data-localisation requirements. Robbin will need to demonstrate that its MCP implementation satisfies those obligations as it scales.
The company has not yet disclosed revenue, customer counts, or the volume of credit decisions processed through the platform. Named customer case studies and independent benchmark data on decisioning accuracy would be the key near-term proof points for enterprise buyers evaluating the product.