An ontology, in Finnlay's words, is "just like a tech word for a queryable graph" of "nodes and edges" — built over time, "non-monotonic" so "the present can affect the past." Fluency's has three layers: defining what work is, knowing who's who across tools, and understanding what's being built.
More from this episode
"A node represents some data and an edge is the relationship." The graph "is being built over time and it's constantly changing depending on new context we get." It's "non-monotonic, which means that the the present can affect the past because sometimes you might get an email or a communication or something changes in your business which affects what we thought was the truth previously."
The three layers: the first "defines what work is"; the second makes sure "we always know it's you" since "to a computer perspective you are different in emails in Slack in the CRM"; the third understands "what is being built through these processes," "whether it's an invoice or a PowerPoint presentation or a unit of code."