AI Middleware Fabric

Our focus is on continuously enhancing the infrastructure of fabric that collects knowledge, understands your data, and provides comprehensive insights into all AI interactions, serving as middleware between an organization’s infrastructure (data, security policies, etc.) and the underlying generative AI framework.

Audit and Governance: We explore innovative ways to ensure that every AI interaction and all associated metadata for future governance and audit tasks is auditable. We do not offer a governance framework, but rather an auditing framework that collects necessary data, enhancing our potential to provide governance capabilities or collaborate with partners specializing in this area.

Data Connectivity and Knowledge Integration: Our R&D work focuses on connecting data and subject matter knowledge to generative AI tools. This includes our ‘know your data’ catalog and knowledge capture mechanisms.

Security Fabric: We invest in researching innovative ways to ensure the security fabric encompasses an organization’s security controls, zero-trust policies, and data policies within the generative AI framework. This is crucial to maintaining data access policies end-to-end.

Tools and Data Integration: We are continuously expanding our suite of tools and data sources to augment the underlying agents. We partner with data providers like D&B to enhance their data sources for agent interactions.

Generative AI Multi-Agent Systems (MAS)

Our current work centers on multi-agent systems rather than zero-shot LLM interactions. We’ve been working with various agent orchestration frameworks and technologies, including AutoGen, CrewAI, devChat, LangGraph, AgencySwarm, and others. Additionally, we’ve been exploring retrieval-augmented generation approaches, leveraging agentic workflows to enhance context for analytical requests, particularly with smaller, air-gapped local models.