Ferris is relevant to the AI agent ecosystem because it provides the essential middle layer between raw enterprise data and agentic action. Their focus on real-time event streaming and microservices allows agents to operate in a reactive, event-driven manner rather than being limited to static data queries. This is the 'nervous system' required for agents that need to respond to market changes or regulatory events as they happen.
Furthermore, Ferris addresses the reliability problem for agents in high-stakes environments. By combining RAG with knowledge graphs, they provide the structured context agents need to minimize hallucinations. For developers building agents for the financial or insurance sectors, Ferris offers the infrastructure to connect those agents to legacy on-premise systems while ensuring every action remains compliant and auditable. They are essentially championing an architecture where agents are a component of a broader, human-orchestrated data mesh.
Ferris emerged from Integration Alpha GmbH, a service provider that began developing data and AI platforms for major clients in 2016. In 2021, the company formalized this development by establishing Ferris Labs AG. This transition reflects a shift from custom consulting to a product-led approach. Today, the organization is split into Ferris Labs AG, which focuses on the core technology, and Ferris Solutions AG, which handles digital solution delivery. Based in Switzerland, the company has built its reputation in the WealthTech sector, serving large financial institutions like UBS, Credit Suisse, and Julius Baer.
The central thesis of the Ferris platform is that 'now' is the only relevant time for data in operational contexts. While many enterprise data stacks rely on batch processing, Ferris prioritizes real-time event streaming and microservices. This focus is designed to move AI from a retrospective reporting tool to an active participant in business processes. By integrating data governance directly into an API-first and event-first architecture, they enable a data mesh approach where information is treated as a product rather than a static resource.
A significant portion of the Ferris value proposition involves legacy application refactoring. In industries like banking, valuable intellectual property is often trapped in decades-old software silos. Ferris uses generative AI co-pilots and its orchestration platform to rewrite and modernize these applications. This process is not just about code migration; it is about breaking up silos and making services reusable across the cloud and on-premise systems. Their anything-as-code methodology ensures that every configuration and data pipeline is versionable and portable, which is a requirement for meeting the auditing standards of the Swiss financial sector.
Ferris approaches generative AI with a skepticism toward unmanaged reasoning. They argue that large language models are mathematical constructs incapable of abstract reasoning, requiring human-designed guardrails. To make GenAI viable for enterprise use, Ferris combines Retrieval-Augmented Generation (RAG) with knowledge graphs. This combination allows models to interact with private, confidential documents while maintaining accuracy and security. Their platform supports local language models and fine-tuning on enterprise data, providing an alternative for organizations that cannot risk sending sensitive data to public cloud AI providers. This security-by-design approach is intended to provide a competitive advantage in sectors where compliance is the primary barrier to AI adoption.
A real-time data and AI orchestration platform for enterprise solutions.
Ferris is hiring.