AI Search Index with large number of fields

[annawiewer]

I’m currently working with a very complex product catalog and extensive requirements around faceting and filtering. As it stands, they will hit the technical limitation with the 1,000 fields per index cap, which makes it difficult to fit everything into a single AI Search index. I’d love to get your insights on how to approach scalable AI Search index design in this kind of scenario. In addition, the product catalog exists in 200 different languages. Specifically, I’m curious about:

• Strategies for working with indices with a lot of fields
• Experience with merging and re-ranking the results in case there are multiple indexes
• Performance implications
• Lessons learned and best practices from similar projects

Thanks!