How do you make centuries of handwritten archives accessible to everyone, regardless of language or ability? In this technical case study, see how Capgemini and Microsoft partnered with the City of Amsterdam to reimagine access to the world’s largest municipal archive – over 50 million digitized documents – using Microsoft Foundry and Azure AI Speech.

We’ll show how Microsoft Foundry provided the backbone for orchestrating modular AI agents, enabling scalable, secure, and maintainable AI powered user interaction. Azure AI Speech technologies powered a conversational interface, allowing users to explore Amsterdam’s history by asking questions in natural language across text, voice, and multiple languages. Advanced OCR and NLP pipelines transformed archaic Dutch manuscripts into searchable, readable content, while speech synthesis and recognition enabled hands-free accessible exploration.

This session will deep dive into the technical architecture: from secure cloud deployment and agent orchestration to integrating Foundry’s agent framework with Azure AI Speech APIs. We’ll share practical lessons on accessibility and scaling conversational AI for sensitive, high-value data—demonstrating how these technologies set a new standard for digital heritage projects.

Key Takeaways:

• Architecting digitization and AI-powered search over vast, complex historical data

• Applying abstraction and modularity with Microsoft Foundry for scalable agent orchestration

• Building accessible, conversational AI experiences using Azure AI Speech

• Real-world outcomes, accessibility breakthroughs, and future directions for digital archives