Embed the world: Multimodal AI for searchable aerial imagery at scale
Amazon Nova Multimodal Embeddings just solved geospatial search at scale. Here's what actually moved the needle.
Why it matters
AWS demonstrates practical multimodal AI deployment for enterprise geospatial search, showing which embedding models and fusion strategies deliver measurable F1 improvements—directly applicable to builders scaling vision-based discovery systems.
The key facts
9 to knowAmazon Nova Multimodal Embeddings achieved highest F1 scores on benchmark queries
System built on Amazon Bedrock and Amazon OpenSearch Serverless
Evaluation methodology based on OpenStreetMap ground truth data
Four experiments compared embedding models, fusion strategies, captioning, and search methods
Work evolved into Vexcel Intelligence product (searchable imagery at scale)
Published Jun 22, 2026
Amazon Nova Multimodal Embeddings delivered highest F1 scores in evaluation
Evolved into Vexcel Intelligence product launch
Focus on design choices for geospatial semantic search optimization
Go to the source
AWS Machine Learning Blogaws.amazon.com
Publisher excerpt: In this post, we walk through the problem space, our architecture on Amazon Bedrock and Amazon OpenSearch Serverless, the evaluation methodology we built on OpenStreetMap ground truth, four experiments that compared embedding models, fusion strategies, captioning, and search methods, and the…
