From Idea to Prototype: Building Compliant AI Solutions for Smart Cities
At the Hackathon Soluciones con WhatsApp para Smart Cities Madrid, AImpacto Rotaract Madrid led a workshop titled “De Idea a Prototipo: ¿Puede tu app incorporar IA, cumplir la normativa europea y escalar desde el primer prototipo?”, delivered by Anna Shatalova (Founder of AImpacto and Compliance Specialist) and Fernando Rangel (AI Specialist and Founder of Kuentia). The session focused on a central challenge often overlooked in early-stage innovation: building AI-enabled solutions that are not only technically functional but also compliant and scalable within the European regulatory framework. Using the case of municipal cultural centers in Madrid (21 in total), we examined how procedural complexity can create unequal access and administrative burden, and how conversational AI, implemented responsibly, can streamline processes without compromising rights.
Participants were guided through the architecture of a functional MVP combining WhatsApp, a structured decision layer, Google Sheets, and OpenAI, emphasizing that the goal is not to build “intelligent” systems for their own sake, but reliable, traceable tools that improve accessibility. A significant portion of the workshop addressed the EU AI Act and the Product Liability Directive, clarifying that regulation focuses on impact, not technology per se, and that systems must be assessed based on the rights and risks they affect.
We introduced the concept of defensive logging — the ability to reconstruct inputs, outputs, and decisions — as essential infrastructure rather than bureaucratic overhead. The key message was clear: if you cannot technically reconstruct a decision, you are not ready to scale. This workshop reflected AImpacto’s core philosophy that innovation without governance is fragile, and governance without innovation is stagnant. Responsible AI begins at the prototype stage, not after deployment.
Disclaimer: redacted with AI.