diff --git a/docs/README.md b/docs/README.md
index 732f671..c6ab6bd 100644
--- a/docs/README.md
+++ b/docs/README.md
@@ -88,6 +88,13 @@ Read the [syntax description](syntax.md) to learn about WireViz' features and ho
 
 See the [tutorial page](../tutorial/readme.md) for sample code, as well as the [example gallery](../examples/readme.md) to see more of what WireViz can do.
 
+### Demo 03
+wireviz -q "design a wire harness that connects two ethernet devices with d38999 connectors on both sides"
+
+Use open AI to design the harness for you. 
+
+Define your openAI API key by setting an environment variable. 
+`OPENAI_KEY={}`
 
 ## Usage
 
diff --git a/src/wireviz/openai.py b/src/wireviz/openai.py
new file mode 100644
index 0000000..c2e08a2
--- /dev/null
+++ b/src/wireviz/openai.py
@@ -0,0 +1,89 @@
+# Hit the chat gpt api to ask for harness design
+
+from openai import OpenAI
+import os
+
+import subprocess
+# set api key to environment variable OPENAI_KEY
+
+import json
+import sys
+
+PROMPT_CONTEXT = '''
+    You are a helpful assistant. 
+    only output a yml file with pinout, notes, pin numbers. Please do not output any information other than the yml. It should follow this format:
+
+connectors:
+  J1: 
+    pinlabels: ["TX+", "TX-", "RX+", "RX-","BI_DD+", "BI_DD-", "BI_DC+", "BI_DC-"]
+    type: "D38999"
+    subtype: receptacle
+  J2:
+    pinlabels: ["TX+", "TX-", "RX+", "RX-","BI_DD+", "BI_DD-", "BI_DC+", "BI_DC-"]
+    type: "D38999"
+    subtype: receptacle
+
+
+cables:
+  W1:
+    wirecount: 8
+    length: 2
+    gauge: 24 AWG
+    show_equiv: true
+    color_code: T568A
+    shield: true # cable shielding included
+    notes: Connect the cable shield to the backshell for EMI grounding
+
+connections:
+  - # Connect twisted pairs for Ethernet
+    - J1: [1-8]
+    - W1: [1-8]
+    - J2: [1-8]
+    
+    '''
+
+# Query the OpenAI API
+def queryGPT(query, model="gpt-4"):
+    """
+    Takes a query and returns the response from the OpenAI API.
+    
+    Parameters:
+        query (str): The input query or prompt.
+        model (str): The model to use for the query (default is "gpt-4").
+    
+    Returns:
+        str: The response from the OpenAI API.
+    """
+    API_KEY= os.getenv("OPENAI_KEY")
+    client = OpenAI(api_key=API_KEY)
+ 
+    query = f"{PROMPT_CONTEXT}\n\n\n{query}"
+    try:
+        # Make an API call
+        response = client.chat.completions.create(model=model,
+        messages=[
+            {"role": "system", "content": PROMPT_CONTEXT},
+            {"role": "user", "content": query}
+        ])
+        # Extract the response content
+        resp =  response.choices[0].message.content.strip()
+        return clean_gpt_response(resp)
+    except Exception as e:
+        return f"An error occurred: {e}"
+    
+
+def clean_gpt_response(response):
+    """
+    Cleans the GPT response by removing any leading or trailing whitespace.
+    
+    Parameters:
+        response (str): The GPT response to clean.
+    
+    Returns:
+        str: The cleaned GPT response.
+    """
+    # remove everything before the word connectors. and remove any trailing ``
+    response = response[response.find("connectors"):]
+    cleaned_output = response.replace("`", "")
+
+    return cleaned_output
\ No newline at end of file
diff --git a/src/wireviz/wireviz.py b/src/wireviz/wireviz.py
index 391b1ac..d4c5d37 100755
--- a/src/wireviz/wireviz.py
+++ b/src/wireviz/wireviz.py
@@ -89,6 +89,7 @@ def parse(
         raise Exception("No output formats or return types specified")
 
     yaml_data, yaml_file = _get_yaml_data_and_path(inp)
+    print(yaml_data)
     if not isinstance(yaml_data, dict):
         raise TypeError(
             f"Expected a dict as top-level YAML input, but got: {type(yaml_data)}"
@@ -380,7 +381,7 @@ def parse(
     if "additional_bom_items" in yaml_data:
         for line in yaml_data["additional_bom_items"]:
             harness.add_bom_item(line)
-
+    print(output_formats)
     if output_formats:
         harness.output(filename=output_file, fmt=output_formats, view=False)
 
diff --git a/src/wireviz/wv_cli.py b/src/wireviz/wv_cli.py
index c83e1cc..2f168ec 100644
--- a/src/wireviz/wv_cli.py
+++ b/src/wireviz/wv_cli.py
@@ -12,6 +12,7 @@ if __name__ == "__main__":
 import wireviz.wireviz as wv
 from wireviz import APP_NAME, __version__
 from wireviz.wv_helper import file_read_text
+from wireviz.openai import queryGPT
 
 format_codes = {
     # "c": "csv",
@@ -71,7 +72,18 @@ epilog += ", ".join([f"{key} ({value.upper()})" for key, value in format_codes.i
     default=False,
     help=f"Output {APP_NAME} version and exit.",
 )
-def wireviz(file, format, prepend, output_dir, output_name, version):
+
+# add an option to input a chatgpt query
+@click.option(
+    "-q",
+    "--query",
+    default=None,
+    type=str,
+    help="Query to input to chatgpt.",
+)
+
+
+def wireviz(file, format, prepend, output_dir, output_name, version, query):
     """
     Parses the provided FILE and generates the specified outputs.
     """
@@ -80,13 +92,24 @@ def wireviz(file, format, prepend, output_dir, output_name, version):
     if version:
         return  # print version number only and exit
 
-    # get list of files
-    try:
-        _ = iter(file)
-    except TypeError:
-        filepaths = [file]
+    # check query
+    if query:
+        gpt_response = queryGPT(query)
+        yml_file= Path("/Users/gouthamsubramanian/harness.yml")
+        with open(yml_file, "w") as file:
+            file.write(gpt_response)
+        print("Response written to", yml_file)
+        filepaths = [yml_file]
     else:
-        filepaths = list(file)
+        try:
+            _ = iter(file)
+        except TypeError:
+            filepaths = [file]
+        else:
+            filepaths = list(file)
+
+
+
 
     # determine output formats
     output_formats = []
@@ -115,7 +138,6 @@ def wireviz(file, format, prepend, output_dir, output_name, version):
     else:
         prepend_input = ""
 
-    # run WireVIz on each input file
     for file in filepaths:
         file = Path(file)
         if not file.exists():