This example shows how to create a semantic document search with chat capability in Rust, using Shuttle and Qdrant - a vector search engine.
- Install Shuttle CLI tools:
cargo install cargo-shuttle
-
Get a Qdrant Account at qdrant.tech
-
Get an OpenAI API key at openai.com
-
Create a
Secrets.toml
file (orSecrets.dev.toml
for local development) with the following contents:
OPENAI_API_KEY = "<ENTER YOUR KEY HERE>"
QDRANT_TOKEN = "<ENTER YOUR KEY HERE>"
QDRANT_URL = "<ENTER YOUR KEY HERE>"
cargo shuttle run
- If you run the service but it only lists the files loaded and doesn't try to embed them, check you have credits/funding in your OpenAI account.
- Make sure your Qdrant URL ends with port
:6334
.