Meet our first model

Text + Time-Series Embeddings

Unified encoding of text and signals in the same vector space.

Flexible Dimensionality

Choose dimensions to balance latency and accuracy for your use case (32–1024).

Drop-in for Text Tasks

Works like any text embedding model for search, clustering, and similarity.

Time-Series Classification & Detection

Embeddings as features for anomaly detection, segmentation, and classification.

Cross-Modal Retrieval

Find similar text or signals across modalities using natural language queries.

Production-Ready Vectors

Stable scales for easy indexing, monitoring, and serving in production.