📈 Vector Search with Qdrant

Now let's add vector storage and search capabilities to the Potato.Examples solution using Qdrant.

---

1. 📦 Install Packages

From the Infrastructure project:

cd Potato.Examples.Infrastructure

dotnet add package Potato.Vectors

dotnet add package Potato.Vectors.Qdrant

---

2. ⚙️ Configure appsettings.json

"PotatoQdrant": {
  "Host": "localhost",
  "Port": 6334
}

---

3. 🛠 Register Services

In Program.cs of your API project:

builder.Services.AddPotatoQdrantVectors(builder.Configuration);

Ensure the collection exists at startup:

app.Lifetime.ApplicationStarted.Register(async () =>
{
    await app.EnsureQdrantCollectionExists();
});

---

4. 🚀 Example Endpoints

app.MapPost("/vectors/add", async (float[] vector, string id, PotatoVectors vec) =>
{
    await vec.Upsert(id, vector).GetOrThrow();
    return Results.Ok();
});

app.MapGet("/vectors/search", async (float[] vector, PotatoVectors vec) =>
{
    var results = await vec.Search(vector).GetOrThrow();
    return Results.Ok(results);
});

---

5. 🐳 Add Qdrant to docker-compose.yml

qdrant:
  image: qdrant/qdrant
  ports:
    - "6333:6333"
    - "6334:6334"

Start it with:

docker compose up -d

---

✅ Summary

Qdrant is now integrated for vector storage and similarity search.

➡️ Next: Potato RAG Pipeline Quickstart Guide