diff --git a/agents.tar b/agents.tar
new file mode 100644
index 0000000..c2c13bf
Binary files /dev/null and b/agents.tar differ
diff --git a/k8s/curaflow-agent/deployment.yaml b/k8s/curaflow-agent/deployment.yaml
index faf34e3..19e814e 100644
--- a/k8s/curaflow-agent/deployment.yaml
+++ b/k8s/curaflow-agent/deployment.yaml
@@ -30,10 +30,10 @@ spec:
           value: "default"
         - name: CURAFLOW_TASK_QUEUE
           value: "curaflow-tasks"
-        - name: OLLAMA_BASE_URL
-          value: "http://ollama.ai-core.svc.cluster.local:11434/v1"
+        - name: ANTHROPIC_API_KEY
+          value: "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"
         - name: LLM_MODEL
-          value: "qwen2.5:3b"
+          value: "claude-3-haiku-20240307"
         resources:
           limits:
             cpu: 500m
diff --git a/requirements.txt b/requirements.txt
index 1b59aee..0b58354 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -2,6 +2,7 @@
 requests
 langchain
 langchain-openai
+langchain-anthropic
 temporalio
 pydantic
 
diff --git a/src/curaflow_agent/activities.py b/src/curaflow_agent/activities.py
index 3a251e4..1aca744 100644
--- a/src/curaflow_agent/activities.py
+++ b/src/curaflow_agent/activities.py
@@ -2,23 +2,20 @@ import os
 import json
 from temporalio import activity
 from langchain_openai import ChatOpenAI
+from langchain_anthropic import ChatAnthropic
 from langchain_core.prompts import PromptTemplate
 from pydantic import BaseModel, Field
 from .models import ClinicalNote, BillingCodesOutput
 from .lab_models import AnomalyReport
 
-# In the cluster, Ollama provides an OpenAI-compatible endpoint.
-# We fetch model/URL from env vars so it's easy to override for testing.
-OLLAMA_BASE_URL = os.environ.get("OLLAMA_BASE_URL", "http://ollama.ai-core.svc.cluster.local:11434/v1")
-LLM_MODEL = os.environ.get("LLM_MODEL", "qwen2.5:3b")
+ANTHROPIC_API_KEY = os.environ.get("ANTHROPIC_API_KEY", "")
+LLM_MODEL = os.environ.get("LLM_MODEL", "claude-3-haiku-20240307")
 
 def get_llm():
-    return ChatOpenAI(
+    return ChatAnthropic(
         model=LLM_MODEL,
-        api_key="ollama", # Ollama doesn't strictly need an API key
-        base_url=OLLAMA_BASE_URL,
+        api_key=ANTHROPIC_API_KEY,
         temperature=0.1,
-        model_kwargs={"response_format": {"type": "json_object"}}
     )
 
 @activity.defn
@@ -187,6 +184,9 @@ async def parse_booking_intent_activity(chat_history: list[dict]) -> dict:
         "Analyze the conversation. Extract any appointment details (department, date, time).\n"
         "If they want to book an appointment but haven't specified department, date, and time, "
         "add those to missing_info and formulate a reply_message asking for them.\n"
+        "CRITICAL INSTRUCTIONS FOR ASKING:\n"
+        "- If the user hasn't specified a department, you MUST list the 4 available departments (Cardiology, General Medicine, Pediatrics, Orthopedics) so they know what to choose. Do not make them guess!\n"
+        "- If they haven't specified a time, suggest common intervals (e.g. 10:00 AM, 11:00 AM, 6:00 PM).\n"
         "If all details are present, set missing_info to empty and formulate a confirmation reply_message.\n"
         "Format output strictly as JSON matching the schema."
     )