Graph Exchange is a quarterly meetup and unconference dedicated to the Graph AI space.

In this instance, we are covering two projects: Semiont and Palefire. We'll also have a discussion, unconference-style, following the talks.

(1) Semiont: A Semantic Knowledge Framework for Human/Agent Co-Creation
Adam Pingel, IBM

Semiont is an open-source semantic knowledge kernel where humans and AI agents co-create shared knowledge as equals — sharing job queues, annotations, and governance rather than bolting AI onto human workflows as an afterthought. Built on W3C Web Annotation and grounded in primary artifacts like PDFs and images, Semiont is designed for high-trust domains that demand provenance. Adam will walk through Semiont's five key flows — Attention, Annotation, Resolution, Context Correlation, and Resource Generation — and demo the system end to end.

Adam Pingel is the Head of Open Tools and Applications for the AI Alliance at IBM. Adam has spent decades at the intersection of software engineering, AI, and knowledge systems. His career includes building AI-powered legal analytics at LexisNexis, where he served as CTO of Global Platforms following the acquisition of Ravel Law, and earlier roles at companies including Excite.com. He holds computer science degrees from UCLA and Stanford, where he studied programming languages, AI, and linguistics.

(2) Pale Fire - Intelligent Knowledge Graph Search System
Slava Tikhonov, Alexy Khrabrov

Pale Fire combines the use of Large Language Models and Knowledge Graphs (semantic artifacts) to support navigation through collections. Situated in the research area of Retrieval Augmented Generation, this specific workflow represents the creation of local and adaptable chatbots. Based on the tool-suite EverythingData at the backend, GhostWriter provides an interface that enables querying and “chatting” with a collection. Applied iteratively, the workflow supports the information needs of researchers when interacting with a collection of papers, whether it be to gain an overview, to learn more about a specific concept and its context, and helps the researcher ultimately to refine their research question in a controlled way.

Pale Fire is developed by Slava Tikhonov, Head of AI and Interoperability at Codata, the European open data consortium.

Codata collaborates with the Community Research Center for Reliable AI at Northeastern Oakland, lead by Alexy Khrabrov, who is presenting this talk.

Pale Fire is now a part of the AgStack, a Linux Foundation project dedicated to Open Source for Agriculture.

(3) Unconference

Following the main talks, we'll have an unconference, where participants can give short talks and we all can discuss the key topics of Knowledge Layer in AI.

Graph Exchange: Context Graph, Annotated Knowledge, and Vector+GraphRAG for More Accurate Q&A