Mozilla, through its subsidiary MZLA Technologies Corporation, has launched Thunderbolt, an open-source AI client designed to give organizations full control over how they build and run AI systems. Unlike proprietary AI platforms, Thunderbolt is self-hostable and customizable, allowing companies to keep their data, infrastructure, and workflows entirely in-house. The platform is positioned as a “sovereign AI” solution—focused on independence, transparency, and freedom from vendor lock-in.
At its core, Thunderbolt acts as an extensible AI workspace where users can interact with AI through chat, search, and research tools while connecting directly to enterprise data systems. Organizations can choose which AI models to run—from commercial providers to open-source or local models—while also automating workflows such as reporting, monitoring, and task execution. The platform is built to operate across all major devices and operating systems, with strong security features including self-hosting, encryption options, and device-level controls.
AI is too important to outsource. With Thunderbolt, we’re giving organizations a sovereign AI client that allows them to decide how AI fits into their workflows – on their infrastructure, with their data, and on their terms.”
Ryan Sipes, CEO of MZLA Technologies Corporation
A key part of Thunderbolt’s offering is its integration with deepset’s Haystack, an open-source AI infrastructure framework. This partnership connects the front-end AI experience with powerful backend capabilities like agents and retrieval-augmented generation (RAG). By linking both layers, organizations can build and operate complete AI systems within a unified architecture, rather than relying on fragmented tools.
This combined stack allows organizations to move beyond isolated AI use cases and toward fully operational AI systems embedded into daily workflows. Thunderbolt handles the user-facing experience, while Haystack manages how AI is built, deployed, and governed. Together, they provide a comprehensive solution for enterprises and public sector organizations seeking greater control over AI systems—from development to deployment.
Thunderbolt is now available via waitlist, with its source code released on GitHub. The launch reflects a growing demand for sovereign AI solutions, where organizations prioritize control, security, and long-term independence over convenience. As AI adoption accelerates, platforms like Thunderbolt aim to redefine how enterprises deploy AI—putting ownership and governance back in their hands
Find out more at
