elastic-opentelemetry-instrumentation-openai: add examples

instrumentation/elastic-opentelemetry-instrumentation-openai/README.md

@@ -23,10 +23,13 @@ pip install elastic-opentelemetry-instrumentation-openai
 This instrumentation supports *zero-code* / *autoinstrumentation*:
 
 ```
-opentelemetry-instrument python use_openai.py
+# "examples/chat.py" is a simple script using the openai client library.
+cd examples
+
+opentelemetry-instrument python chat.py
 
 # You can record more information about prompts as log events by enabling content capture.
-OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT=true opentelemetry-instrument python use_openai.py
+OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT=true opentelemetry-instrument python chat.py
 ```
 
 Or manual instrumentation:

instrumentation/elastic-opentelemetry-instrumentation-openai/examples/README.md

@@ -0,0 +1,71 @@
+# OpenAI Zero-Code Instrumentation Examples
+
+This is an example of how to instrument OpenAI calls with zero code changes,
+using `opentelemetry-instrument` included in the Elastic Distribution of
+OpenTelemetry Python ([EDOT Python][edot-python]).
+
+When OpenAI examples run, they export traces, metrics and logs to an OTLP
+compatible endpoint. Traces and metrics include details such as the model used
+and the duration of the LLM request. In the case of chat, Logs capture the
+request and the generated response. The combination of these provide a
+comprehensive view of the performance and behavior of your OpenAI usage.
+
+## Install
+
+First, set up a Python virtual environment like this:
+```bash
+python3 -m venv .venv
+source .venv/bin/activate
+pip install -r requirements.txt
+```
+
+Next, install [EDOT Python][edot-python] and dotenv which is a portable way to
+load environment variables.
+```bash
+pip install "python-dotenv[cli]" elastic-opentelemetry
+```
+
+Finally, run `edot-bootstrap` which analyzes the code to add relevant
+instrumentation, to record traces, metrics and logs.
+```bash
+edot-bootstrap --action=install
+```
+
+## Configure
+
+Minimally, update the [.env](.env) file with your `OPENAI_API_KEY`.
+
+An OTLP compatible endpoint should be listening for traces, metrics and logs on
+`http://localhost:4317`. If not, update `OTEL_EXPORTER_OTLP_ENDPOINT` as well.
+
+## Run
+
+There are two examples, and they run the same way:
+
+### Chat
+
+[chat.py](chat.py) asks the LLM a geography question and prints the response.
+
+Run it like this:
+```bash
+dotenv run -- opentelemetry-instrument python chat.py
+```
+
+You should see something like "Atlantic Ocean" unless your LLM hallucinates!
+
+### Embeddings
+
+
+[embeddings.py](embeddings.py) creates in-memory VectorDB embeddings about
+Elastic products. Then, it searches for one similar to a question.
+
+Run it like this:
+```bash
+dotenv run -- opentelemetry-instrument python embeddings.py
+```
+
+You should see something like "Connectors" unless your LLM hallucinates!
+
+---
+
+[edot-python]: https://github.com/elastic/elastic-otel-python/blob/main/docs/get-started.md

instrumentation/elastic-opentelemetry-instrumentation-openai/examples/chat.py

@@ -0,0 +1,23 @@
+import os
+
+import openai
+
+CHAT_MODEL = os.environ.get("CHAT_MODEL", "gpt-4o-mini")
+
+
+def main():
+    client = openai.Client()
+
+    messages = [
+        {
+            "role": "user",
+            "content": "Answer in up to 3 words: Which ocean contains the falkland islands?",
+        }
+    ]
+
+    chat_completion = client.chat.completions.create(model=CHAT_MODEL, messages=messages)
+    print(chat_completion.choices[0].message.content)
+
+
+if __name__ == "__main__":
+    main()

instrumentation/elastic-opentelemetry-instrumentation-openai/examples/embeddings.py

@@ -0,0 +1,49 @@
+import os
+
+import numpy as np
+import openai
+
+EMBEDDINGS_MODEL = os.environ.get("EMBEDDINGS_MODEL", "text-embedding-3-small")
+
+
+def main():
+    client = openai.Client()
+
+    products = [
+        "Search: Ingest your data, and explore Elastic's machine learning and retrieval augmented generation (RAG) capabilities."
+        "Observability: Unify your logs, metrics, traces, and profiling at scale in a single platform.",
+        "Security: Protect, investigate, and respond to cyber threats with AI-driven security analytics."
+        "Elasticsearch: Distributed, RESTful search and analytics.",
+        "Kibana: Visualize your data. Navigate the Stack.",
+        "Beats: Collect, parse, and ship in a lightweight fashion.",
+        "Connectors: Connect popular databases, file systems, collaboration tools, and more.",
+        "Logstash: Ingest, transform, enrich, and output.",
+    ]
+
+    # Generate embeddings for each product. Keep them in an array instead of a vector DB.
+    product_embeddings = []
+    for product in products:
+        product_embeddings.append(create_embedding(client, product))
+
+    query_embedding = create_embedding(client, "What can help me connect to OpenAI?")
+
+    # Calculate cosine similarity between the query and document embeddings
+    similarities = []
+    for product_embedding in product_embeddings:
+        similarity = np.dot(query_embedding, product_embedding) / (
+            np.linalg.norm(query_embedding) * np.linalg.norm(product_embedding)
+        )
+        similarities.append(similarity)
+
+    # Get the index of the most similar document
+    most_similar_index = np.argmax(similarities)
+
+    print(products[most_similar_index])
+
+
+def create_embedding(client, text):
+    return client.embeddings.create(input=[text], model=EMBEDDINGS_MODEL, encoding_format="float").data[0].embedding
+
+
+if __name__ == "__main__":
+    main()

instrumentation/elastic-opentelemetry-instrumentation-openai/examples/requirements.txt

@@ -0,0 +1,2 @@
+openai~=1.57.2
+numpy~=2.2.0

scripts/license_headers_check.sh

@@ -17,7 +17,7 @@
 
 if [ $# -eq 0 ]
 then
-   FILES=$(find . \( -name "*.py" -o -name "*.c" -o -name "*.sh" \) -size +1c -not -path "./dist/*" -not -path "./build/*" -not -path "./venv*/*")
+   FILES=$(find . \( -name "*.py" -o -name "*.c" -o -name "*.sh" \) -size +1c -not -path "./dist/*" -not -path "./build/*" -not -path "./venv*/*" -not -path "*/examples/*")
 else
     FILES=$@
 fi