Einstein Copilot vs Complex Queries
Summary
The article evaluates the improved capabilities of Salesforce Einstein Copilot's latest Atlas Reasoning Engine when handling complex multi-step queries involving related Salesforce objects. It shows how Copilot can correctly interpret and execute complex queries about Opportunities, Products, and related criteria, but still struggles with certain keywords that break the logic chain. The author tests different query phrasings and highlights the current limitations, emphasizing that custom Apex actions remain necessary until the AI assistant matures further. Salesforce teams can experiment with phrasing their Copilot queries or maintain custom code to handle sophisticated filtering needs in CRM contexts.
Takeaways
- Einstein Copilot handles multi-object relationship queries better but still has edge case failures.
- Custom Apex actions remain important for complex queries until AI matures further.
- Rephrasing natural language requests can influence Copilot query success.
- Copilot’s logic can confuse object relationship IDs causing query failures.
- Not all Atlas Reasoning Engine features are available until early 2025.
Image created by GPT-4o based in a prompt by Bob Buzzard Introduction Now that Agentforce (for Service at least) is GA we have access to the latest Atlas Reasoning Engine . That's the theory at least - I haven't seen a way to find out what version is in use, which models it has access to etc, but that doesn't worry me too much as I can't change or influence it anyway. Anecdotally I do feel that the ability to handle requests with more than one step has improved over the last few months, but this sounded like a step change - Chain of Thought reasoning and an iterative approach to finding the best response! My focus continues to be on Copilot (aka Agentforce for CRM, or whatever name it's going by this week), but I'm writing rather more custom actions than I'd like.