Skip to main content
This guide walks you through migrating your collections and vectors from Qdrant to Actian VectorAI DB. Follow the steps below to export, transform, and import your data.
1

Export data from Qdrant

Use the Qdrant REST API or Python SDK to extract your vectors, payloads, and IDs. Export to a portable format such as JSON, CSV, or NumPy arrays.
2

Prepare data for VectorAI DB

Map your exported fields to the VectorAI DB point structure: unique IDs, vector embeddings, and optional JSON metadata payloads. Ensure vector dimensions match the target collection configuration and that data types are compatible.
3

Install VectorAI DB

If you have not already set up VectorAI DB, follow the Docker installation guide to deploy it locally or on your server.
4

Ingest data into VectorAI DB

Use the Python SDK to create a collection with matching dimensions and distance metric, then insert your transformed vectors and metadata. Refer to the Python SDK quickstart for code examples.
5

Validate migration

Confirm the migration succeeded by comparing record counts, vector dimensions, and sample search results between your Qdrant instance and VectorAI DB. Check that metadata fields are intact and that similarity queries return consistent rankings.
For large datasets, use batch inserts and parallel processing to improve ingestion performance. Start with smaller subsets to validate your pipeline before running the full migration.