Put the data you can't expose to work.
Match records with organizations that can't share theirs, and run AI on your own sensitive data, while the data stays encrypted.

Turn data that can't be shared into decisions no one could make alone.
DoubleBlind matches records across organizations' separate datasets in real time, while every record stays encrypted. Each organization keeps its data in its own environment, under its own keys. Only the permitted answer comes back.
Query goes out encrypted.
The query is encrypted with fully homomorphic encryption, so it can be searched while it stays encrypted.
Each organization searches its own records.
The search runs on encrypted data, in each organization's own environment, and the query stays hidden from them.
Only the permitted answer comes back.
A match, a count, or which fields agree, whatever the check is set to disclose. The records stay with their owners.
Enabling previously impossible workflows
Put your most sensitive data to work in AI, without exposing it.
Vector search on encrypted data. Self-hosted, with your keys.
- Disk-native scaleTiered by design: quantized vectors in memory, full vectors on local disk or S3-compatible object storage.
- Filter-first retrievalA global metadata index filters first, then searches the matching set, so recall holds as filters stack up.
- Encrypted in useSearch runs over ciphertext. Only the final candidates are decrypted, in your environment, with your KMS.
- One-container deployOne container image in your environment, persisting to the object storage you already run.
Encrypted. Faster than most plaintext.
On wiki-all-1M at 99% recall, CyborgDB beats Weaviate, Milvus, Elasticsearch, Qdrant, LanceDB and pgvector — encrypted.
| Database | Recall (%) | QPS |
|---|---|---|
| CyborgDB (encrypted) | 72.8 | 792 |
| CyborgDB (encrypted) | 76.3 | 744 |
| CyborgDB (encrypted) | 81.3 | 672 |
| CyborgDB (encrypted) | 85.9 | 638 |
| CyborgDB (encrypted) | 88.8 | 567 |
| CyborgDB (encrypted) | 91.8 | 499 |
| CyborgDB (encrypted) | 94.0 | 433 |
| CyborgDB (encrypted) | 96.2 | 351 |
| CyborgDB (encrypted) | 97.3 | 321 |
| CyborgDB (encrypted) | 98.2 | 266 |
| CyborgDB (encrypted) | 98.6 | 245 |
| CyborgDB (encrypted) | 99.0 | 220 |
| CyborgDB (encrypted) | 99.4 | 182 |
| CyborgDB (encrypted) | 99.8 | 151 |
| Qdrant | 89.5 | 113 |
| Qdrant | 95.2 | 89 |
| Qdrant | 97.8 | 68 |
| Qdrant | 99.1 | 45 |
| Qdrant | 99.4 | 35 |
| Qdrant | 99.7 | 27 |
| Qdrant | 99.9 | 17 |
| Weaviate | 77.2 | 840 |
| Weaviate | 83.4 | 737 |
| Weaviate | 86.9 | 715 |
| Weaviate | 91.3 | 600 |
| Weaviate | 93.3 | 533 |
| Weaviate | 95.9 | 430 |
| Weaviate | 97.5 | 349 |
| Weaviate | 98.6 | 266 |
| Weaviate | 98.9 | 224 |
| Weaviate | 99.4 | 176 |
| Weaviate | 99.8 | 108 |
| Milvus | 92.6 | 66 |
| Milvus | 96.7 | 61 |
| Milvus | 98.3 | 55 |
| Milvus | 99.1 | 47 |
| Milvus | 99.5 | 41 |
| Milvus | 99.7 | 29 |
| Elasticsearch | 81.4 | 303 |
| Elasticsearch | 83.4 | 293 |
| Elasticsearch | 89.6 | 278 |
| Elasticsearch | 90.8 | 276 |
| Elasticsearch | 94.6 | 241 |
| Elasticsearch | 95.2 | 240 |
| Elasticsearch | 96.6 | 218 |
| Elasticsearch | 97.5 | 198 |
| Elasticsearch | 97.8 | 184 |
| Elasticsearch | 98.7 | 161 |
| Elasticsearch | 99.2 | 131 |
| Elasticsearch | 99.2 | 129 |
| Elasticsearch | 99.4 | 120 |
| Elasticsearch | 99.4 | 117 |
| pgvector | 81.4 | 835 |
| pgvector | 89.2 | 497 |
| pgvector | 94.1 | 402 |
| pgvector | 96.7 | 283 |
| pgvector | 97.6 | 190 |
| pgvector | 98.5 | 130 |
| pgvector | 99.1 | 82 |
| pgvector | 99.5 | 47 |
| pgvector | 99.6 | 38 |
| LanceDB | 94.2 | 282 |
| LanceDB | 97.5 | 174 |
| LanceDB | 99.0 | 101 |
| LanceDB | 99.3 | 90 |
Single-threaded runs on the ann-benchmarks harness, c8g.4xlarge · May 2026.
What teams build on CyborgDB when the data can't leave their control.
Secure RAG,
at NVIDIA scale.
The Cyborg–NVIDIA Enterprise RAG Blueprint is a reference architecture for deploying secure AI applications at NVIDIA scale. Cyborg isone of eight companies NVIDIA invited to author an Enterprise Blueprint.
NVIDIA NIMs handle the inference. CyborgDB handlesencrypted-in-use retrieval across multimodal corpora.
- 15× fastermultimodal PDF extraction
- 50% fewerincorrect answers
- Encryptedsearch and storage

Your most valuable data is the data you can't expose.
Cyborg puts it to work.
Match records with organizations that can't share theirs using DoubleBlind, or run AI on your own sensitive data with CyborgDB.