RAG & Embedding

	modified:   docs/genai/04_how_to_guides/02_embeddings.mdx
	modified:   docs/genai/04_how_to_guides/03_retrieval_augmented_generation.mdx

Author: s-sajid-ali

docs/genai/04_how_to_guides/02_embeddings.mdx

@@ -4,9 +4,9 @@ While Decoder-only LLMs gained massive popularity via their usage in chatbots, E
 
 ```mermaid
 flowchart LR;
-    A["String <br> "GenAI can be used for research""]
-    B["Encoder-only LLM"]
-    C["Vector embedding <br> [0.052587852, 0.094195396, 0.24439038, 0.104940414, ...]"]
+    A["natual language text string <br> *GenAI can be used for research*"]
+    B["encoder-only LLM"]
+    C["vector embedding <br> [0.052587852, 0.094195396, 0.24439038, 0.104940414, ...]"]
     A-- "Input" -->B;
     B-- "Output" -->C;
 ```

docs/genai/04_how_to_guides/03_retrieval_augmented_generation.mdx

@@ -5,3 +5,25 @@ Large Language Models only know about the data they were trained upon and do not
 - newer knowledge past the cutoff date (i.e. the date at which data collection was frozen)
 
 To get around this issue, one of the most popular techniques is Retrieval-augmented generation.
+
+
+```mermaid
+flowchart TB;
+    A["Natual language prompt <br> *Can GenAI be used for research?*"]
+    B["Vector embedding <br> [0.052587852, 0.094195396, 0.24439038, 0.104940414, ...]"]
+    C["Vector Database <br> embedding1 <br> embedding2 <br> embedding3 <br> ... "]
+    D["Text with embeddings similar to the prompt"]
+    E["Original prompt with added context"]
+    F["Response from LLM using context"]
+    subgraph Retreival
+    direction LR
+    A-- "Embedding" -->B;
+    B-- "Look for similar embeddings" -->C;
+    C-- "Generate context" -->D;
+    end
+    subgraph Augmented Generation
+    direction LR
+    D-- "Expand Prompt" -->E;
+    E-- "LLM" -->F;
+    end
+```