About This Research
This master's thesis investigates how local, self-hosted multi-agent systems can handle complex enterprise tasks while maintaining strict privacy and data governance. Organizations increasingly need AI systems that preserve data residency, minimize external data transfer, and provide transparent auditability—requirements that traditional cloud-based approaches struggle to satisfy.
The research explores using local open-weight language models coordinated through workflow orchestration to create responsible AI systems with verifiable execution, consent management, and comprehensive audit trails. This work is conducted in collaboration with Siili Solutions under the TechFinland100 grant program.
Current Status
Try It Out
I'm building a prototype chatbot that demonstrates some of the concepts from this research. You can interact with it using the chat button in the bottom-right corner to see how multi-agent systems might coordinate to handle complex enterprise tasks.
The chatbot is currently in development and uses mock responses for demonstration purposes.