Integrations
InertialAI provides integration packages that connect the inertial-embed-alpha embedding model directly to popular data platforms. These packages implement each platform's native interfaces, so you can add InertialAI's multi-modal embeddings to your existing infrastructure without writing custom HTTP clients or embedding pipelines.
Available Integrations
Vector Databases
| Integration | Package | Status | Description |
|---|---|---|---|
| Chroma | inertialai-chroma · GitHub | Available | Drop-in EmbeddingFunction for ChromaDB. Supports text-only, time-series-only, and multi-modal inputs. |
| Pinecone | inertialai-pinecone | Coming soon | Managed vector database for high-scale production similarity search. |
| Weaviate | inertialai-weaviate | Coming soon | Open-source vector database with a GraphQL query interface and built-in ML model integrations. |
| Qdrant | inertialai-qdrant | Coming soon | Open-source vector search engine with rich filtering support and on-premise deployment options. |
| Milvus / Zilliz | inertialai-milvus | Coming soon | Cloud-native, billion-scale embedding retrieval. |
Time-Series Databases
| Integration | Package | Status | Description |
|---|---|---|---|
| InfluxDB | inertialai-influxdb | Coming soon | Persist raw time-series data in InfluxDB while simultaneously embedding it with InertialAI for downstream AI tasks — a natural fit given InertialAI's focus on sensor data. |
Why Use an Integration Package?
Integration packages handle the entire bridge between InertialAI's embeddings API and your storage layer:
- Drop-in compatibility — implement each platform's native embedding interface, so existing tooling works without modification.
- Multi-modal input support — embed plain text, raw time-series, or text paired with time-series data in the same collection.
- Secure credential handling — API keys are read from environment variables and are never persisted to disk.
- Request lifecycle management — authentication, payload formatting, response parsing, and error propagation are handled for you.
Getting Started
The Chroma integration is available today. If you do not have a Chroma instance running yet, see the Chroma Docker deployment guide to get one up in minutes.
To generate embeddings directly via InertialAI's API without an integration package, see: