- MUST READ: Submission requirements
- Entry form template
Core Power Delta Task Force Omega
Best tool supporting Sitecore Development (not built on Sitecore)
Core Power is a streamlined, one-stop platform for Sitecore learning and development, built to support developers at every stage of their journey. We created Core Power to simplify how developers learn the Sitecore ecosystem, since the official documentation—while comprehensive—can often feel overwhelming, especially for newcomers. Its largely flat structure makes it difficult to understand how concepts relate to one another, which topics are foundational, and which areas are most important for certifications or real-world development. Core Power addresses this by organizing knowledge into clear, logical learning paths that group related concepts together and highlight their importance within the broader platform. To further personalize the experience, the platform uses AI agents to intelligently suggest learning paths based on a developer’s goals and experience level, along with a chatbot that helps answer questions, explain concepts, and guide users through the learning journey.
Visit https://two026-core-power-delta-task-force-omega.onrender.com to enjoy the experience. This a standalone app accessible in a browser. There are no additional installation instructions.
The app homepage is a diagram of core competencies for the default Sitecore product, Sitecore AI. The user sees these sized base on their relative prevalance/emphasis across documentation and community data (forums, blogs, etc). Homepage
Here, the user can:
-
select a different product to explore using a dropdown as shown here Product Selection
-
select a topic by clicking it Select topic
-
viewing topic summary/related resources and optionally start testing knowledge with a quiz upon selection Select topic
-
take Agent generated quiz to test platform knowledge and recieve feedback on answers Take quiz Quiz feedback
-
interact with Agent via chat for conversational platform support Agent chat
Additional implementation details are available in the following README files:
Serialization Agent Service README
A note on agent hallucinations - To combat agent hallucinations, our team instrumented each of our agents (chat and quiz) with LangSmith. LangSmith is a tool that allows us to capture telemetry on tool calls, token consumption, and I/O within each agent. We plan to use this tracing alongside our established RAG capabilities to build multi-agent validation systems that ensure generated content, such as recommendations and links to blogs, forums, and docs, is accurate.
Subset of documentation - For this iteration, we are only concerned with a subset of the full Sitecore documentation. We felt this would allow us to establish a process for capturing, refining, and leveraging Sitecore information that could be repeated across different contexts.
Volatile vector storage - One of our priorities for the agentic chat was speed. Because of this, we opted to leverage an in-memory vector database called ChromaDB. This cuts down on latency; however, it required us to orchestrate seeding vectors on API startup. This is not something we planned to bring into production, but a trade-off we accepted given the MVP nature of this build.