Knowledge graphs give AI real-time context, Intuit says

The conversation at the Neo4j GraphTalk event highlights a shift in how enterprises approach AI. As language models become commodities, Intuit's experience suggests the differentiator is not the model itself but the context layer underneath it. Cloes's comments indicate that the discipline of building a graph — connecting data in relevant ways — creates a reusable foundation that outlasts any single AI vendor.

Intuit's reported journey from days to seconds in analysis illustrates a concrete operational win. Security operations, compliance, and real-time context were the initial drivers, but the same graph layer now powers AI tooling. This positions the graph not as a niche technology but as core infrastructure for AI readiness, where the value lies in democratized, queryable data.

Cloes's indifference to which AI model wins — "ChatGPT today, Amazon Bedrock tomorrow" — underscores a pragmatic strategy. By decoupling the data context from the model, enterprises can avoid lock-in and remain flexible. This suggests that the long-term winners may be those who invest in the connective tissue of their data, rather than in any single AI platform.