Submission by Juan Ortiz

I built a voice-based AI agent for Monet, a Colombian fintech focused on microcredit access, designed to answer user questions and guide them through understanding financial services in a simple and trustworthy way. The problem it solves is the lack of accessible, clear, and real-time financial guidance for users who are unfamiliar with fintech products. Many potential users don’t fully understand how services like microcredit work, which creates friction, distrust, and drop-off in the exploration phase. This agent reduces that barrier by providing instant, conversational, and easy-to-understand answers. The solution uses ElevenLabs as the core conversational and voice layer, enabling natural, human-like interactions in Spanish. The agent leverages ElevenLabs’ multilingual voice model to deliver a friendly and վստահable experience through voice and WhatsApp channels. Additionally, it integrates Firecrawl (this week’s sponsor) as a tool to dynamically scrape and scroll Monet’s website in real time. This allows the agent to ground its responses in up-to-date company information, ensuring accuracy without relying solely on static prompts. Together, this creates a scalable, voice-first assistant that combines conversational AI, real-time data retrieval, and fintech education to improve user trust and accessibility.