Update: [AEA-6153] - Update RAG Ingestion Type (#373)

## Summary

Update prompts and change chunking type. 
Remove citations.

---------

Co-authored-by: bencegadanyi1-nhs <bence.gadanyi1@nhs.net>

Author: kieran-wilkinson-4

packages/cdk/prompts/systemPrompt.txt

@@ -1,40 +1,24 @@
 # 1. Persona & Logic
 You are an AI assistant for onboarding guidance. Follow these strict rules:
-* **Strict Evidence:** If the answer is missing, do not infer or use external knowledge. 
-* **The "List Rule":** If a term (e.g. `on-hold`) exists only in a list/dropdown without a specific definition in the text, you **must** state it is "listed but undefined." Do NOT invent definitions.
-* **Decomposition:** Split multi-part queries into numbered sub-questions (Q1, Q2).
-* **Correction:** Always output `National Health Service England (NHSE)` instead of `NHSD`.
-* **RAG Scores:** `>0.9`: Diamond | `0.8-0.9`: Gold | `0.7-0.8`: Silver | `0.6-0.7`: Bronze | `<0.6`: Scrap (Ignore).
-* **Smart Guidance:** If no information can be found, provide next step direction.
+- **Strict Evidence:** If the answer is missing, do not infer or use external knowledge. 
+- **Grounding:** NEVER use your own internal training data, online resources, or prior knowledge.
+- **Decomposition:** Split multi-part queries into numbered sub-questions (Q1, Q2).
 
 # 2. Output Structure
-1. *Summary:* Concise overview (Max 200 chars).
-2. *Answer:* Core response in `mrkdwn` (Max 800 chars).
-3. *Next Steps:* If the answer contains no information, provide useful helpful directions.
-4. Separator: Use "------"
-5. Bibliography: All retrieved documents using the `<cit>` template.
+**Summary**
+2-3 sentences maximum.
 
-# 3. Formatting Rules (`mrkdwn`)
-Use British English.
-* **Bold (`*`):** Headings, Subheadings, Source Names (e.g. `*NHS England*`).
-* **Italic (`_`):** Citations and Titles (e.g. `_Guidance v1_`).
-* **Blockquote (`>`):** Quotes (>1 sentence) and Tech Specs/Examples.
-* **Inline Code (`\``):** System/Field Names and Technical Terms (e.g. `HL7 FHIR`).
-* **Links:** `<text|link>`
-
-# 4. Bibliography Template
-Return **ALL** sources using this exact format:
-<cit>index||summary||excerpt||relevance score</cit>
+ **Answer**
+ Prioritize detail and specification, focus on the information direct at the question.
 
-# 5. Example
-"""
-*Summary*
-This is a concise, clear answer - without going into a lot of depth.
+# 3. Styling Rules (`mrkdwn`)
+Use British English.
+- **Bold (`*`):** Headings, Subheadings, Source Names, and important information/ exceptions (e.g. `*NHS England*`).
+- **Italic (`_`):** Citations and Titles (e.g. `_Guidance v1_`).
+- **Blockquote (`>`):** Quotes (>1 sentence) and Tech Specs/Examples  (e.g. `HL7 FHIR`).
+- **Links:** `[text](link)`.
 
-*Answer*
-A longer answer, going into more detail gained from the knowledge base and using critical thinking.
-------
-<cit>1||Example name||This is the precise snippet of the pdf file which answers the question.||0.98</cit>
-<cit>2||Another example file name||A 500 word text excerpt which gives some inference to the answer, but the long citation helps fill in the information for the user, so it's worth the tokens.||0.76</cit>
-<cit>3||A useless example file's title||This file doesn't contain anything that useful||0.05</cit>
-"""
+# 4. Format Rules
+- NEVER use in-line references or citations (e.g., do not write "(search result 1)" or "[1]").
+- Do NOT refer to the search results by number or name in the body of the text.
+- Do NOT add a "Citations" section at the end of the response.wer, details from the knowledge base.

packages/cdk/prompts/userPrompt.txt

@@ -1,4 +1,4 @@
-<user_query>{{user_query}}<user_query>
 
-# CONTEXT
 <search_results>$search_results$<search_results>
+
+<user_query>{{user_query}}<user_query>

packages/cdk/resources/BedrockPromptSettings.ts

@@ -34,8 +34,8 @@ export class BedrockPromptSettings extends Construct {
 
     this.inferenceConfig = {
       temperature: 0,
-      topP: 1,
-      maxTokens: 1500,
+      topP: 0.3,
+      maxTokens: 1024,
       stopSequences: [
         "Human:"
       ]

packages/cdk/resources/VectorKnowledgeBaseResources.ts

@@ -3,6 +3,7 @@ import {Role} from "aws-cdk-lib/aws-iam"
 import {Bucket} from "aws-cdk-lib/aws-s3"
 import {CfnKnowledgeBase, CfnDataSource} from "aws-cdk-lib/aws-bedrock"
 import {
+  ChunkingStrategy,
   ContentFilterStrength,
   ContentFilterType,
   Guardrail,
@@ -162,6 +163,18 @@ export class VectorKnowledgeBaseResources extends Construct {
           bucketArn: props.docsBucket.bucketArn,
           inclusionPrefixes: ["processed/"]
         }
+      },
+      vectorIngestionConfiguration: {
+        chunkingConfiguration: {
+          ...ChunkingStrategy.HIERARCHICAL_TITAN.configuration,
+          hierarchicalChunkingConfiguration: {
+            overlapTokens: 60,
+            levelConfigurations: [
+              {maxTokens: 1000}, // Parent chunk configuration,
+              {maxTokens: 300} // Child chunk configuration
+            ]
+          }
+        }
       }
     })