Context engines and knowledge graphs: Sigma’s AI team explains

In this video, Marwan and Sri explain context engines and how we’re using them at Sigma to make our AI experiences more accurate and reliable while cutting latency and cost. We use natural language question-answering agents as our example, but much of this applies to agents more broadly. What we cover: - What an agent is and the tools it uses - Three types of context: data, institutional knowledge, and agent interaction history - How a context engine replaces a sprawl of tool calls with a few efficient ones - How context resolves ambiguity in “revenue in my region” - The Sigma knowledge graph and its lift from 59% to 82% accuracy on our internal evaluation We’re building the Sigma context engine to power Sigma Assistant and agents built on the Sigma platform for every Sigma customer. New to Sigma? Request a demo: https://www.sigmacomputing.com/go/request-a-demo?utm_medium=organic-social&utm_source=youtube&utm_campaign=request-a-demo&utm_content=request-a-demo&utm_term=request-a-demo Free trial: https://www.sigmacomputing.com/go/free-trial?utm_medium=organic-social&utm_source=youtube&utm_campaign=free-trial&utm_content=free-trial&utm_term=free-trial