feat: add AWS Bedrock function calling implementation

plugins/aws/README.md

@@ -28,9 +28,9 @@ agent = Agent(
 
 The full example is available in example/aws_qwen_example.py
 
-### Realtime Text/Image Usage
+### Realtime Audio Usage
 
-Nova sonic audio realtime STS is also supported:
+Nova Sonic audio realtime STS is also supported:
 
 ```python    
 agent = Agent(
@@ -41,6 +41,62 @@ agent = Agent(
 )
 ```
 
+## Function Calling
+
+### Standard LLM (aws.LLM)
+
+The standard LLM implementation **fully supports** function calling. Register functions using the `@llm.register_function` decorator:
+
+```python
+from vision_agents.plugins import aws
+
+llm = aws.LLM(
+    model="qwen.qwen3-32b-v1:0",
+    region_name="us-east-1"
+)
+
+@llm.register_function(
+    name="get_weather",
+    description="Get the current weather for a given city"
+)
+def get_weather(city: str) -> dict:
+    """Get weather information for a city."""
+    return {
+        "city": city,
+        "temperature": 72,
+        "condition": "Sunny"
+    }
+```
+
+### Realtime (aws.Realtime)
+
+The Realtime implementation **fully supports** function calling with AWS Nova Sonic. Register functions using the `@llm.register_function` decorator:
+
+```python
+from vision_agents.plugins import aws
+
+llm = aws.Realtime(
+    model="amazon.nova-sonic-v1:0",
+    region_name="us-east-1"
+)
+
+@llm.register_function(
+    name="get_weather",
+    description="Get the current weather for a given city"
+)
+def get_weather(city: str) -> dict:
+    """Get weather information for a city."""
+    return {
+        "city": city,
+        "temperature": 72,
+        "condition": "Sunny"
+    }
+
+# The function will be automatically called when the model decides to use it
+```
+
+See `example/aws_realtime_function_calling_example.py` for a complete example.
+
 ## Running the examples
 
 Create a `.env` file, or cp .env.example to .env and fill in

plugins/aws/example/aws_llm_function_calling_example.py

@@ -0,0 +1,89 @@
+import asyncio
+import logging
+from uuid import uuid4
+
+from dotenv import load_dotenv
+
+from vision_agents.core import User
+from vision_agents.core.agents import Agent
+from vision_agents.plugins import aws, getstream, cartesia, deepgram, smart_turn
+
+load_dotenv()
+
+logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s [call_id=%(call_id)s] %(name)s: %(message)s")
+logger = logging.getLogger(__name__)
+
+
+async def start_agent() -> None:
+    agent = Agent(
+        edge=getstream.Edge(),
+        agent_user=User(name="Weather Bot"),
+        instructions="You are a helpful weather bot. Use the provided tools to answer questions.",
+        llm=aws.LLM(
+            model="anthropic.claude-3-sonnet-20240229-v1:0",
+            region_name="us-east-1"
+
+        ),
+        tts=cartesia.TTS(),
+        stt=deepgram.STT(),
+        # turn_detection=smart_turn.TurnDetection(buffer_duration=2.0, confidence_threshold=0.5),
+    )
+
+    # Register custom functions
+    @agent.llm.register_function(
+        name="get_weather",
+        description="Get the current weather for a given city"
+    )
+    def get_weather(city: str) -> dict:
+        """Get weather information for a city."""
+        logger.info(f"Tool: get_weather called for city: {city}")
+        if city.lower() == "boulder":
+            return {"city": city, "temperature": 72, "condition": "Sunny"}
+        return {"city": city, "temperature": "unknown", "condition": "unknown"}
+
+    @agent.llm.register_function(
+        name="calculate",
+        description="Performs a mathematical calculation"
+    )
+    def calculate(expression: str) -> dict:
+        """Performs a mathematical calculation."""
+        logger.info(f"Tool: calculate called with expression: {expression}")
+        try:
+            result = eval(expression)  # DANGER: In a real app, use a safer math evaluator!
+            return {"expression": expression, "result": result}
+        except Exception as e:
+            return {"expression": expression, "error": str(e)}
+
+    await agent.create_user()
+
+    call = agent.edge.client.video.call("default", str(uuid4()))
+    await agent.edge.open_demo(call)
+
+    with await agent.join(call):
+        # Give the agent a moment to connect
+        await asyncio.sleep(5)
+
+        # Test function calling with weather
+        logger.info("Testing weather function...")
+        await agent.llm.simple_response(
+            text="What's the weather like in Boulder? Please use the get_weather function."
+        )
+
+        await asyncio.sleep(5)
+
+        # Test function calling with calculation
+        logger.info("Testing calculation function...")
+        await agent.llm.simple_response(
+            text="Can you calculate 25 multiplied by 4 using the calculate function?"
+        )
+
+        await asyncio.sleep(5)
+
+        # Wait a bit before finishing
+        await asyncio.sleep(5)
+        await agent.finish()
+
+
+if __name__ == "__main__":
+    asyncio.run(start_agent())
+

plugins/aws/example/aws_realtime_function_calling_example.py

@@ -0,0 +1,141 @@
+import asyncio
+import logging
+from pathlib import Path
+from uuid import uuid4
+
+from dotenv import load_dotenv
+
+from vision_agents.core import User
+from vision_agents.core.agents import Agent
+from vision_agents.plugins import aws, getstream
+
+load_dotenv()
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s %(levelname)s [call_id=%(call_id)s] %(name)s: %(message)s"
+)
+logger = logging.getLogger(__name__)
+
+
+async def start_agent() -> None:
+    """Example demonstrating AWS Bedrock realtime with function calling.
+
+    This example creates an agent that can call custom functions to get
+    weather information and perform calculations.
+    """
+
+    # Create the agent with AWS Bedrock Realtime
+    agent = Agent(
+        edge=getstream.Edge(),
+        agent_user=User(name="Weather Assistant AI"),
+        instructions="""You are a helpful weather assistant. When users ask about weather,
+        use the get_weather function to fetch current conditions. You can also help with
+        simple calculations using the calculate function.""",
+        llm=aws.Realtime(
+            model="amazon.nova-sonic-v1:0",
+            region_name="us-east-1",
+            voice_id="matthew"
+        ),
+    )
+
+    # Register custom functions that the LLM can call
+    @agent.llm.register_function(
+        name="get_weather",
+        description="Get the current weather for a given city"
+    )
+    def get_weather(city: str) -> dict:
+        """Get weather information for a city.
+
+        Args:
+            city: The name of the city
+
+        Returns:
+            Weather information including temperature and conditions
+        """
+        # This is a mock implementation - in production you'd call a real weather API
+        weather_data = {
+            "Boulder": {"temp": 72, "condition": "Sunny", "humidity": 30},
+            "Seattle": {"temp": 58, "condition": "Rainy", "humidity": 85},
+            "Miami": {"temp": 85, "condition": "Partly Cloudy", "humidity": 70},
+        }
+
+        city_weather = weather_data.get(city, {"temp": 70, "condition": "Unknown", "humidity": 50})
+        return {
+            "city": city,
+            "temperature": city_weather["temp"],
+            "condition": city_weather["condition"],
+            "humidity": city_weather["humidity"],
+            "unit": "Fahrenheit"
+        }
+
+    @agent.llm.register_function(
+        name="calculate",
+        description="Perform a mathematical calculation"
+    )
+    def calculate(operation: str, a: float, b: float) -> dict:
+        """Perform a calculation.
+
+        Args:
+            operation: The operation to perform (add, subtract, multiply, divide)
+            a: First number
+            b: Second number
+
+        Returns:
+            Result of the calculation
+        """
+        operations = {
+            "add": lambda x, y: x + y,
+            "subtract": lambda x, y: x - y,
+            "multiply": lambda x, y: x * y,
+            "divide": lambda x, y: x / y if y != 0 else None,
+        }
+
+        if operation not in operations:
+            return {"error": f"Unknown operation: {operation}"}
+
+        result = operations[operation](a, b)
+        if result is None:
+            return {"error": "Cannot divide by zero"}
+
+        return {
+            "operation": operation,
+            "a": a,
+            "b": b,
+            "result": result
+        }
+
+    # Create and start the agent
+    await agent.create_user()
+
+    call = agent.edge.client.video.call("default", str(uuid4()))
+    await agent.edge.open_demo(call)
+
+    with await agent.join(call):
+        # Give the agent a moment to connect
+        await asyncio.sleep(5)
+
+        # Test function calling with weather
+        logger.info("Testing weather function...")
+        await agent.llm.simple_response(
+            text="What's the weather like in Boulder? Please use the get_weather function."
+        )
+
+        await asyncio.sleep(5)
+
+        # Test function calling with calculation
+        logger.info("Testing calculation function...")
+        await agent.llm.simple_response(
+            text="Can you calculate 25 multiplied by 4 using the calculate function?"
+        )
+
+        await asyncio.sleep(5)
+
+        # Wait a bit before finishing
+        await asyncio.sleep(5)
+        await agent.finish()
+
+
+if __name__ == "__main__":
+    asyncio.run(start_agent())
+

plugins/aws/example/pyproject.toml

@@ -1,22 +1,22 @@
 [project]
-name = "gemini-live-realtime-example"
+name = "aws-bedrock-realtime-example"
 version = "0.0.0"
-requires-python = ">=3.10"
+requires-python = ">=3.12"
 
 # put only what this example needs
 dependencies = [
   "python-dotenv>=1.0",
-  "vision-agents-plugins-gemini",
+  "vision-agents-plugins-aws",
   "vision-agents-plugins-getstream",
   "vision-agents",
-  "google-genai>=1.33.0",
+  "boto3>=1.26.0",
   "opentelemetry-exporter-otlp>=1.37.0",
   "opentelemetry-exporter-prometheus>=0.58b0",
   "prometheus-client>=0.23.1",
   "opentelemetry-sdk>=1.37.0",
 ]
 
 [tool.uv.sources]
-"vision-agents-plugins-getstream" = {path = "../../../plugins/getstream", editable=true}
-"vision-agents-plugins-gemini" = {path = "../../../plugins/gemini", editable=true}
+"vision-agents-plugins-getstream" = {path = "../../getstream", editable=true}
+"vision-agents-plugins-aws" = {path = "..", editable=true}
 "vision-agents" = {path = "../../../agents-core", editable=true}