Thanks for the detailed explanation — really appreciate the clarity around native waterfall enrichments.
In our case, Clay’s native enrichments alone won’t work for the full use case. We’re dealing with a large, custom enrichment workflow where enrichment needs to be AI-driven, conditional, and orchestrated externally based on context, logic, and downstream system requirements.
Specifically:
Enrichment must be triggered selectively and dynamically, not uniformly across rows
The logic depends on Claude-driven reasoning (classification, prioritization, normalization) before enrichment is invoked
We need tighter control over when, why, and how enrichment happens across multiple datasets and systems
Clay is still a core enrichment engine for us, but it needs to operate as part of a Claude ↔️ Clay integration rather than a standalone native-only flow.
Given this constraint, we’d love guidance on:
The best-supported way to integrate Claude with Clay at scale
Whether webhooks or HTTP/API integrations are preferred for this kind of AI-orchestrated workflow
Any patterns or limitations we should be aware of when Clay is used as an enrichment service rather than the primary orchestrator
