NotifAI Performance Improvements

.gitignore

@@ -242,3 +242,5 @@ src/backend/app/.env
 src/backend/bedrock_alerts/lambda_build
 src/backend/bedrock_alerts/lambda_function.zip
 src/backend/bedrock_alerts/lambda_function.zip
+src/backend/bedrock_alerts/evaluations_alerts.zip
+plugin-cache

docker-compose.yaml

@@ -0,0 +1,93 @@
+x-aws-environment: &aws-environment
+  AWS_DEFAULT_REGION: eu-west-2
+  AWS_REGION: eu-west-2
+  AWS_SECRET_ACCESS_KEY: ${AWS_SECRET_ACCESS_KEY}
+  AWS_ACCESS_KEY_ID: ${AWS_ACCESS_KEY_ID}
+  AWS_SESSION_TOKEN: ${AWS_SESSION_TOKEN}
+
+services:
+  frontend:
+    build:
+      context: src/frontend/notifai-uploader
+    ports:
+      - "3000:80"
+    depends_on:
+      - backend
+      - bedrock-prompt-messager
+    environment:
+      <<: *aws-environment
+      REACT_APP_BACKEND_API_BASE_URL: "http://localhost:8080"
+      REACT_APP_COGNITO_ID: "4kgkrqjc0770v6slu27pjuagb1"
+      REACT_APP_COGNITO_USER_POOL_ID: "eu-west-2_JuQWLuCTr"
+      REACT_APP_API_GATEWAY: "http://localhost:8081/call-llm"
+    healthcheck:
+      test: ["CMD", "curl", "http://localhost:80/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 30s
+
+  backend:
+    build:
+      context: src/backend/app
+    ports:
+      - "8080:8080"
+    entrypoint:
+      [
+        "fastapi",
+        "dev",
+        "/app/main.py",
+        "--port",
+        "8080",
+        "--host",
+        "0.0.0.0",
+        "--reload",
+      ]
+    volumes:
+      - "${PWD}/src/backend/app:/app"
+    environment:
+      <<: *aws-environment
+      COGNITO_REGION: "eu-west-2"
+      COGNITO_USER_POOL_ID: "eu-west-2_JuQWLuCTr"
+      COGNITO_APP_CLIENT_ID: "4kgkrqjc0770v6slu27pjuagb1"
+      S3_LLM_LOGS_BUCKET: "nhs-notifai-shared"
+      S3_LLM_LOGS_DIRECTORY: "logs/"
+      S3_LLM_LOGS_BUCKET_ACCOUNT_ID: "496395772806"
+    healthcheck:
+      test:
+        [
+          "CMD",
+          "python3",
+          "-c",
+          "import urllib.request; urllib.request.urlopen('http://localhost:8080/health')",
+        ]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 30s
+
+  bedrock-prompt-messager:
+    build:
+      context: src/backend/bedrock-prompt-messager
+      dockerfile: Dockerfile
+    ports:
+      - "8081:8080"
+    volumes:
+      - "${PWD}/src/backend/bedrock-prompt-messager:/var/task"
+    environment:
+      <<: *aws-environment
+      env_model_id: "amazon.nova-lite-v1:0"
+      env_temperature: "0.1"
+      env_max_tokens: "10000"
+      env_top_p: "0.5"
+      env_logging_s3_bucket: "nhs-notifai-shared"
+      env_logging_s3_key_prefix: "logs/"
+      env_guardrail_arn: "arn:aws:bedrock:eu-west-2:496395772806:guardrail/7z5kf9qg62is"
+      env_guardrail_version: "1"
+      env_logging_s3_account_id: "496395772806"
+    healthcheck:
+      test: ["CMD", "curl", "http://localhost:8080/health"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+      start_period: 30s

infrastructure/terraform/components/notifyai/bedrock.tf

@@ -100,7 +100,7 @@ data "aws_iam_policy_document" "bedrock_access_s3" {
       aws_s3_bucket.evaluation_programatic_results.arn,
       "${aws_s3_bucket.evaluation_programatic_results.arn}/*",
       "arn:aws:bedrock:*::foundation-model/*",
-      "arn:aws:bedrock:${var.region}:${var.aws_account_id}:inference-profile/eu.amazon.nova-pro-v1:0",
+      "arn:aws:bedrock:${var.region}:${var.aws_account_id}:inference-profile/eu.amazon.nova-lite-v1:0",
       "arn:aws:bedrock:${var.region}:${var.aws_account_id}:evaluation-job/*",
       "arn:aws:bedrock:${var.region}::prompt/*",
       "arn:aws:bedrock:*:${var.aws_account_id}:inference-profile/*",

infrastructure/terraform/components/notifyai/eventbridge.tf

@@ -12,6 +12,8 @@ module "eventbridge" {
       arn                 = aws_lambda_function.bedrock_evaluations.arn
       input               = jsonencode({ "job" : "cron-by-rate" })
       role_arn            = aws_iam_role.eventbridge_scheduler_role.arn
+
+      end_date = "2025-12-16T00:00:00Z"
     }
   }
 }

infrastructure/terraform/components/notifyai/lambda.tf

@@ -67,7 +67,7 @@ data "aws_iam_policy_document" "bedrock_access" {
       "${aws_s3_bucket.lambda_prompt_logging_s3_bucket.arn}/${local.s3_lambda_logging_key}*",
       "arn:aws:bedrock:${var.region}:${var.aws_account_id}:guardrail/*",
       "arn:aws:bedrock:${var.region}:${var.aws_account_id}:inference-profile/eu.amazon.nova-pro-v1:*",
-      "arn:aws:bedrock:${var.region}::foundation-model/amazon.nova-pro-v1:0",
+      "arn:aws:bedrock:${var.region}::foundation-model/amazon.nova-lite-v1:0",
       "arn:aws:logs:${var.region}:${var.aws_account_id}:log-group:/aws/lambda/${local.lambda_name}:*"
     ]
   }

infrastructure/terraform/etc/env_eu-west-2_dev1.tfvars

@@ -5,9 +5,9 @@ prompt-max-tokens-to-sample = 5000
 prompt-temperature          = 0.1
 prompt-top-p                = 0.5
 
-prompt-model                          = "amazon.nova-pro-v1:0"
-evaluation-evaluator-model-identifier = "amazon.nova-pro-v1:0"
-evaluation-inference-model-identifier = "amazon.nova-pro-v1:0"
+prompt-model                          = "amazon.nova-lite-v1:0"
+evaluation-evaluator-model-identifier = "amazon.nova-lite-v1:0"
+evaluation-inference-model-identifier = "amazon.nova-lite-v1:0"
 evaluation-schedule-days              = 30
 
 log_retention_in_days = 3

infrastructure/terraform/etc/env_eu-west-2_dev2.tfvars

@@ -5,9 +5,9 @@ prompt-max-tokens-to-sample = 5000
 prompt-temperature          = 0.1
 prompt-top-p                = 0.5
 
-prompt-model                          = "amazon.nova-pro-v1:0"
-evaluation-evaluator-model-identifier = "amazon.nova-pro-v1:0"
-evaluation-inference-model-identifier = "amazon.nova-pro-v1:0"
+prompt-model                          = "amazon.nova-lite-v1:0"
+evaluation-evaluator-model-identifier = "amazon.nova-lite-v1:0"
+evaluation-inference-model-identifier = "amazon.nova-lite-v1:0"
 evaluation-schedule-days              = 30
 
 log_retention_in_days = 3

infrastructure/terraform/etc/env_eu-west-2_test.tfvars

@@ -5,7 +5,7 @@ prompt-max-tokens-to-sample = 5000
 prompt-temperature          = 0.1
 prompt-top-p                = 0.5
 
-prompt-model                          = "amazon.nova-pro-v1:0"
-evaluation-evaluator-model-identifier = "amazon.nova-pro-v1:0"
-evaluation-inference-model-identifier = "amazon.nova-pro-v1:0"
+prompt-model                          = "amazon.nova-lite-v1:0"
+evaluation-evaluator-model-identifier = "amazon.nova-lite-v1:0"
+evaluation-inference-model-identifier = "amazon.nova-lite-v1:0"
 evaluation-schedule-days              = 3

infrastructure/terraform/etc/group_nhs-notify-poc001.tfvars

@@ -4,7 +4,7 @@ aws_account_id = "767397886959"
 region         = "eu-west-2"
 
 # Generics
-log_retention_in_days = 10
+log_retention_in_days = 14
 kms_deletion_window   = 15
 
 budget_amount          = 300

scripts/config/sonar-scanner.properties

@@ -5,7 +5,7 @@ sonar.qualitygate.wait=true
 sonar.sourceEncoding=UTF-8
 sonar.sources=src/backend/app, src/backend/bedrock-prompt-messager, src/frontend/notifai-uploader
 sonar.tests=src/backend/app/tests, src/backend/bedrock-prompt-messager/tests/
-sonar.exclusions=src/**/tests/**
+sonar.exclusions=src/**/tests/**,src/backend/bedrock-prompt-messager/api_wrapper.py,src/backend/bedrock-prompt-messager/Dockerfile
 sonar.python.coverage.reportPaths=coverage.xml
 #sonar.[javascript|typescript].lcov.reportPaths=.coverage/lcov.info
-sonar.coverage.exclusions=scripts/**/*,**/core/config.py,**/core/constants.py,**/frontend/**/*
+sonar.coverage.exclusions=scripts/**/*,**/core/config.py,**/core/constants.py,**/frontend/**/*,src/backend/bedrock-prompt-messager/api_wrapper.py

src/backend/app/Dockerfile

@@ -2,8 +2,6 @@ FROM python:slim
 
 WORKDIR /app
 
-RUN apt-get update && apt-get install -y pandoc && rm -rf /var/lib/apt/lists/*
-
 COPY requirements.txt /requirements.txt
 
 RUN pip install --no-cache-dir --upgrade -r /requirements.txt
@@ -12,4 +10,4 @@ COPY . .
 
 EXPOSE 8080
 
-CMD ["fastapi", "start", "app/main.py", "--port", "8080"]
+CMD ["fastapi", "run", "/app/main.py", "--port", "8080"]

src/backend/app/core/auth.py

@@ -140,6 +140,9 @@ async def dispatch(self, request: Request, call_next):
         if request.method == "OPTIONS":
             return await call_next(request)
 
+        if request.url.path in ["/health"]:
+            return await call_next(request)
+
         try:
             payload = _verify_request_token(request, self.authenticator)
             request.state.user = payload

src/backend/app/main.py

@@ -6,7 +6,7 @@
     CognitoAuthenticator,
     NotFoundExceptionHandler,
 )
-from app.routers import convert, s3
+from app.routers import s3
 from dotenv import load_dotenv
 
 load_dotenv()
@@ -16,18 +16,22 @@
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
-authenticator = CognitoAuthenticator()
-
-app.add_middleware(AuthMiddleware, authenticator=authenticator)
-app.add_exception_handler(404, NotFoundExceptionHandler(authenticator))
-
 app.add_middleware(
     CORSMiddleware,
-    allow_origin_regex=r"https://[a-z0-9]+\.[a-z0-9\-]+\.awsapprunner\.com$",  # Regex for App Runner URLs only
+    # Regex for App Runner and local dev URLs only
+    allow_origin_regex=r"https://[a-z0-9]+\.[a-z0-9\-]+\.awsapprunner\.com|.*localhost.*|127\.0\.0\.1$",
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
 )
 
-app.include_router(convert.router)
+authenticator = CognitoAuthenticator()
+app.add_middleware(AuthMiddleware, authenticator=authenticator)
+app.add_exception_handler(404, NotFoundExceptionHandler(authenticator))
+
 app.include_router(s3.router)
+
+
+@app.get("/health")
+async def health():
+    return {"detail": "ok"}, 200

src/backend/app/requirements.txt

@@ -1,9 +1,7 @@
 fastapi[standard]==0.115.2
 uvicorn==0.32.0
 python-multipart==0.0.18
-boto3
-pandoc
-pypdf
+boto3[crt]
 pytest
 pyjwt[crypto]
 python-jose