Community Call: Making your AI think before doing

Query Plans for Accurate and Steerable AI on Data
Traditional methods of building AI applications on data such as Retrieval Augmented Generation (RAG) and tool calling reduce your AI’s intelligence. They utilize the same pre-defined workflows for any user query to achieve an outcome. User queries are dynamic, and your AI system’s approach to data retrieval, analysis, and action should be dynamic too. In this call, we’ll demonstrate how query planning allows you to build AI systems and agents that separate planning and execution to flexibly adapt to any user query, resulting in accurate and transparent responses to any data.

PromptQL vs. reasoning models
Would you trust AI to make critical business decisions? During this call, we’ll benchmark PromptQL against the latest reasoning models like OpenAI's o1, o3-mini, and Claude 3.5 Sonnet with tool calling. We’ll show how PromptQL transforms natural language into precise operations by decoupling query planning from execution, ensuring full visibility, steerability, accuracy, and repeatability. Join us as we demonstrate how this method outperforms the latest reasoning models, building the trust you need to make high-stakes business decisions.

The PromptQL natural language API
Explore the new PromptQL Natural Language API, a powerful chat API that lets you query in natural language and receive real-time, streaming responses. Discover how this API enables interactive application development with seamless message exchanges and artifact management to elevate your AI apps.

Web connectors for PromptQL
We have a bunch of new web search connectors for PromptQL, integrating Perplexity, Brave Search, or DuckDuckGo. See how these connectors enable your natural language queries to pull in real-time web data, helping you quickly enrich your private data with up-to-date public information.

DDN Updates

• RESTified Endpoints (Demo)
• Comment & Collab feature