Power your AI Chatbot

with 2500+ Pre-Built Actions

Transform your AI chatbot into a powerful tool that not only converses but also executes tasks , workflows and actions — all with just three lines of code.

The API/SDK to enable your AI-ChatBot with actions

‣ Integrate over 2,500 simple or complex single or mult-step actions into your AI agent or chatbot with a few lines of code.

‣ Easy to use API/SDK

‣ Ability to create your own simple and complex single/multi-step actions

‣ Realtime deliverability + webhook for 2-way communication

‣ Enterprise-grade infrastructure backed by SOC2 Type II and HIPAA/PIPEDA Security & Compliance

def get_response():
    """Get the response from the AI and execute the tool call if necessary"""
    response = call_chatgpt(messages, open_ai_client)

    # Verify that the response has a tool call, then execute the tool call
    tool_response = wrk_gpt_client.execute_tool(response)
    if tool_response[0]:

        messages.append(response.choices[0].message)
        messages.append({
            "role": "tool",
            "content": json.dumps(tool_response[0]),
            "tool_call_id": response.choices[0].message.tool_calls[0].id
        })

        # Start a processing thread to monitor the tool that was called
        if tool_response[1]:
            tool_thread = threading.Thread(
                target=monitor_tool, 
                args=(tool_response[1],)
            )
            tool_thread.start()

        response = call_chatgpt(messages, open_ai_client)

    messages.append({"role": "assistant", "content": response.choices[0].message.content})
    
    return response.choices[0].message.content

The API/SDK to enable your AI-ChatBot with actions

‣ Integrate over 2,500 simple or complex single or mult-step actions into your AI agent or chatbot with a few lines of code.

‣ Easy to use API/SDK

‣ Ability to create your own simple and complex single/multi-step actions

‣ Realtime deliverability + webhook for 2-way communication

‣ Enterprise-grade infrastructure backed by SOC2 Type II and HIPAA/PIPEDA Security & Compliance

def get_response():
    """Get the response from the AI and execute the tool call if necessary"""
    response = call_chatgpt(messages, open_ai_client)

    # Verify that the response has a tool call, then execute the tool call
    tool_response = wrk_gpt_client.execute_tool(response)
    if tool_response[0]:

        messages.append(response.choices[0].message)
        messages.append({
            "role": "tool",
            "content": json.dumps(tool_response[0]),
            "tool_call_id": response.choices[0].message.tool_calls[0].id
        })

        # Start a processing thread to monitor the tool that was called
        if tool_response[1]:
            tool_thread = threading.Thread(
                target=monitor_tool, 
                args=(tool_response[1],)
            )
            tool_thread.start()

        response = call_chatgpt(messages, open_ai_client)

    messages.append({"role": "assistant", "content": response.choices[0].message.content})
    
    return response.choices[0].message.content

The API to enable your AI-ChatBot with actions

Integrate over 2,500 simple or complex single or mult-step actions into your AI agent or chatbot with a few lines of code.

‣ Easy to use SDK

‣ Ability to create your own complex multi-step actions

‣ Realtime deliverability + webhook for 2-way communication

‣ Enterprise-grade infrastructure backed by SOC2 Type II Security & Compliance

def get_response():
    """Get the response from the AI and execute the tool call if necessary"""
    response = call_chatgpt(messages, open_ai_client)

    # Verify that the response has a tool call, then execute the tool call
    tool_response = wrk_gpt_client.execute_tool(response)
    if tool_response[0]:

        messages.append(response.choices[0].message)
        messages.append({
            "role": "tool",
            "content": json.dumps(tool_response[0]),
            "tool_call_id": response.choices[0].message.tool_calls[0].id
        })

        # Start a processing thread to monitor the tool that was called
        if tool_response[1]:
            tool_thread = threading.Thread(
                target=monitor_tool, 
                args=(tool_response[1],)
            )
            tool_thread.start()

        response = call_chatgpt(messages, open_ai_client)

    messages.append({"role": "assistant", "content": response.choices[0].message.content})
    
    return response.choices[0].message.content

Start Automating with Wrk

Kickstart your automation journey with the Wrk all-in-one automation platform