Generative Search for your Enterprise
No Embeddings. No VectorDB

Go beyond RAG with ThirdAI NeuralDB to build & deploy generative search on your own documents.

What can you do with Generative Search

Index and Train on 1000s of documents on your laptop.

Or infrastructure of your choice

pip3 install thirdai
db = ndb.NeuralDB(user_id="my_user")

pdf_files = ['data/sample_nda.pdf']
source_ids = db.insert(pdf_files, train=True)
search_results = db.search(
    query="what is the termination period",
    top_k=2)

How it Works

Sentiment Analysis
Method RoBERTa (fine tuned for sentiment) ThirdAI Bolt
Accuracy
83.02%
93%
Training Time
40 hours on GPU
20 min on laptop CPU
Inference latency (ms)
46
1
SciFact Benchmark
Method T5-Large ThirdAI UDT
Precision@1
39
58
Recall@100
82
90
Information Retrieval on MSMarco
Method ColBERT V2 ThirdAI BOLT
Latency (ms)
721
100
Recall
.965
.962

Unlock the Power of LLMs
at a Fraction of the Cost

The ThirdAI engine makes it easy to build and deploy billion parameter models on just CPUs.

No configs. No GPUs. No latency.