Why should you care?

Regulated workloads need vectors inside customer network; Qdrant Docker on Azure VM satisfies data residency.

See it live — copy this example

Paste into an ASP.NET Core 8+ / AgentNest project, then run with dotnet run (set your API keys in user-secrets).

// AgentNest.Infrastructure/QdrantVectorStore.cs
public sealed class QdrantVectorStore(QdrantClient client)
{
    const string Collection = "agentnest_kb";
    public async Task EnsureCollectionAsync(ulong dim, CancellationToken ct)
    {
        if (!(await client.ListCollectionsAsync(ct)).Contains(Collection))
            await client.CreateCollectionAsync(Collection, new VectorParams { Size = dim, Distance = Distance.Cosine }, cancellationToken: ct);
    }
    public async Task UpsertAsync(string tenantId, string id, float[] vector, string text, CancellationToken ct) =>
        await client.UpsertAsync(Collection, [new PointStruct
        {
            Id = id, Vectors = vector,
            Payload = { ["tenant"] = tenantId, ["text"] = text }
        }], cancellationToken: ct);
}

What happened?

• QdrantVectorStore creates cosine collection once and upserts points with tenant payload filters for hospital KB chunks.
• Follow the steps below — typing the code yourself is the fastest way to learn.

Practice next

1. Run Qdrant via docker run -p 6333:6333 qdrant/qdrant.
2. dotnet add package Qdrant.Client.
3. Filter searches with tenant match in Qdrant filter API.
4. Enable quantization for large hospital corpora.
5. Add HNSW ef parameter tuning in collection config.

Remember

Qdrant suits self-hosted AgentNest RAG in customer VPCs. Use payload filters for tenant isolation. Create collection with correct dimension at startup.

On-prem hospital RAG

Hospital IT mandates clinical vectors never leave their Azure VNet.

Outcome: Qdrant cluster inside VNet powers AgentNest assistant with local embeddings API.