diff --git a/src/Templates/Boilerplate/Bit.Boilerplate/.docs/20- .NET Aspire.md b/src/Templates/Boilerplate/Bit.Boilerplate/.docs/20- .NET Aspire.md
index 650f3f11be..a6fb010de2 100644
--- a/src/Templates/Boilerplate/Bit.Boilerplate/.docs/20- .NET Aspire.md	
+++ b/src/Templates/Boilerplate/Bit.Boilerplate/.docs/20- .NET Aspire.md	
@@ -333,6 +333,70 @@ This ensures **consistency** between development and production environments!
 
 ---
 
+## ☁️ Going Production: Switching to Azure Managed Services
+
+While local containers (Docker) are perfect for development, for production, you could switch to fully managed Azure services to ensure scalability, security, and high availability.
+
+.NET Aspire makes this transition seamless by swapping the hosting resources in your `AppHost/Program.cs` without changing your application code.
+
+### 1. Prerequisite Packages
+
+Ensure you have the Azure integration packages installed in your **AppHost** project:
+
+```xml
+<PackageReference Include="Aspire.Hosting.Azure.Sql" />
+<PackageReference Include="Aspire.Hosting.Azure.Redis" />
+<PackageReference Include="Aspire.Hosting.Azure.PostgreSQL" />
+
+```
+
+### 2. Updating AppHost (`Program.cs`)
+
+Replace your local container definitions with Azure resources.
+
+#### 🛢️ Azure SQL Database
+
+Instead of `AddSqlServer` (container), use `AddAzureSqlServer`:
+
+```csharp
+// ❌ Local Container
+// var sqlServer = builder.AddSqlServer("sqlserver");
+
+// ✅ Azure Managed Service
+var sqlServer = builder.AddAzureSqlServer("sqlserver")
+    .AddDatabase("sqldb");
+
+```
+
+#### ⚡ Azure Cache for Redis
+
+Instead of `AddRedis` (container), use `AddAzureRedis`:
+
+```csharp
+// ❌ Local Container
+// var redis = builder.AddRedis("redis");
+
+// ✅ Azure Managed Service
+var redis = builder.AddAzureRedis("redis");
+
+```
+
+#### 🐘 Azure Database for PostgreSQL (Flexible Server)
+
+To leverage high-performance vector search in production, use Azure PostgreSQL Flexible Server.
+
+```csharp
+// ❌ Local Container
+// var postgres = builder.AddPostgres("postgres");
+
+// ✅ Azure Managed Service
+var postgres = builder.AddAzurePostgresFlexibleServer("postgres")
+    .AddDatabase("pgdb");
+
+```
+
+---
+
 ## Additional Resources
 
 - 📚 **Official Documentation**: https://aspire.dev
diff --git a/src/Templates/Boilerplate/Bit.Boilerplate/.docs/25- RAG - Semantic Search with Vector Embeddings (Advanced).md b/src/Templates/Boilerplate/Bit.Boilerplate/.docs/25- RAG - Semantic Search with Vector Embeddings (Advanced).md
index 2515858915..cf8479c373 100644
--- a/src/Templates/Boilerplate/Bit.Boilerplate/.docs/25- RAG - Semantic Search with Vector Embeddings (Advanced).md	
+++ b/src/Templates/Boilerplate/Bit.Boilerplate/.docs/25- RAG - Semantic Search with Vector Embeddings (Advanced).md	
@@ -350,5 +350,91 @@ A hybrid approach can offer a balance between speed and accuracy. You can first
 
 ---
 
+## 7. Performance Optimization: Azure DiskANN index & Reranking
+
+When moving to production with large datasets (e.g., millions of vectors) in **SQL Server** or **PostgreSQL**, it is fully recommended to use the **DiskANN** index. DiskANN provides high-performance, disk-based approximate nearest neighbor (ANN) search.
+
+However, to utilize DiskANN efficiently and maintain high accuracy (recall), you should modify your query strategy. Instead of a simple "Order By & Take", use a **two-step Reranking** approach.
+
+### Steps to enable DiskANN
+
+#### Step A: Enable Extensions in `DbContext`
+
+In your `AppDbContext.cs` (Server.Api project), ensure the extensions are registered:
+Example instructions provided for PostgreSQL
+
+```csharp
+protected override void OnModelCreating(ModelBuilder modelBuilder)
+{    
+    // Enable pgvector extension for vector operations
+    modelBuilder.HasPostgresExtension("vector");
+    
+    // Enable Azure's high-performance DiskANN extension
+    modelBuilder.HasPostgresExtension("pg_diskann"); 
+}
+```
+
+#### Step B: Configure the Index (`EntityTypeConfiguration`)
+
+In your Entity Configuration file (e.g., `ProductConfiguration.cs`), define the DiskANN index. This tells EF Core to create the specialized index:
+
+```csharp
+public void Configure(EntityTypeBuilder<Product> builder)
+{
+    // Define the vector column
+    builder.Property(p => p.Embedding)
+           .HasColumnType("vector(384)"); // Adjust size based on your embedding model
+
+    // Define the DiskANN Index
+    builder.HasIndex(p => p.Embedding)
+           .HasMethod("diskann") // Explicitly use Azure's DiskANN
+           .HasOperators("vector_cosine_ops") // Use cosine similarity
+           .HasStorageParameter("product_quantized", true); // Enable `Product Quantization` to leverage high-dimensional support
+}
+```
+
+### The Query Strategy
+
+1. **Fetch Candidates**: Request more results than you need (e.g., 50) using the approximate index.
+2. **Rerank**: Re-sort those candidates by exact distance and return the top results (e.g., 10).
+
+https://learn.microsoft.com/en-us/azure/postgresql/extensions/how-to-use-pgdiskann#improve-accuracy-when-using-pq-with-vector-reranking
+
+Instead of the standard query, refactor your LINQ query to use a subquery structure. This encourages the database engine to use the index for coarse filtering before performing fine-grained sorting.
+
+**Standard Query (Less Optimized for DiskANN):**
+
+```csharp
+// Simple scan - might be slow or less accurate with heavily compressed indexes
+var value = new Pgvector.Vector(embeddedSearchQuery.Vector);
+return dbContext.Products
+    .Where(p => p.Embedding!.CosineDistance(value!) < DISTANCE_THRESHOLD)
+    .OrderBy(p => p.Embedding!.CosineDistance(value!))
+    .Take(10);
+```
+
+**Recommended Reranking Query:**
+
+```csharp
+var value = new Pgvector.Vector(embeddedSearchQuery.Vector);
+return dbContext.Products
+    // This would require the rest of the filters to be applied here instead of the method's returned IQueryable by the caller. For example Price > X etc.
+    .Where(p => p.Embedding!.CosineDistance(value!) < DISTANCE_THRESHOLD)
+    .OrderBy(p => p.Embedding!.CosineDistance(value!))
+    .Take(CANDIDATE_COUNT) // Step 1: Approximate Search (fetch more candidates for better accuracy). This is especially important when using DiskANN indexes which may sacrifice some accuracy for speed
+    .OrderBy(p => p.Embedding!.CosineDistance(value!))
+    .Take(FINAL_RESULT_COUNT); // Step 2: Reranking (Refine to top results with exact distances)
+```
+
+**Important**: The database execution plan should reflect the re-ranking strategy, ensuring that the DiskANN index is utilized effectively.
+
+**Why this matters:**
+
+* **DiskANN** uses compression (quantization) to be fast.
+* By fetching a larger "candidate list" (50) first, you compensate for potential approximation errors.
+* Sorting the small list (50 items) by exact distance ensures your final Top 10 are highly accurate.
+
+---
+
 Happy coding! 🚀