Updating OpenAI embedding example to use gateways to work better with… (

#2)

… docker run

examples/applications/compute-openai-embeddings/README.md

@@ -0,0 +1,25 @@
+# Computing text embeddings with Open AI
+
+This sample application shows how to use Open AI to compute text embeddings by calling the API.
+
+## Configure you OpenAI API Key
+
+Export to the ENV the access key to OpenAI
+
+```
+export OPEN_AI_ACCESS_KEY=...
+```
+
+## Deploy the LangStream application
+
+```
+./bin/langstream docker run test -app examples/applications/compute-openai-embeddings -s examples/secrets/secrets.yaml
+```
+
+## Talk with the Chat bot using the CLI
+Since the application opens a gateway, we can use the gateway API to send and consume messages.
+
+```
+./bin/langstream gateway chat test -cg output -pg input -p sessionId=$(uuidgen)
+```
+

examples/applications/compute-openai-embeddings/gateways.yaml

@@ -0,0 +1,21 @@
+gateways:
+  - id: "input"
+    type: produce
+    topic: "input-topic"
+    parameters:
+      - sessionId
+    produceOptions:
+      headers:
+        - key: langstream-client-session-id
+          valueFromParameters: sessionId
+
+  - id: "output"
+    type: consume
+    topic: "output-topic"
+    parameters:
+      - sessionId
+    consumeOptions:
+      filters:
+        headers:
+          - key: langstream-client-session-id
+            valueFromParameters: sessionId

examples/applications/compute-openai-embeddings/pipeline.yaml

@@ -25,15 +25,19 @@ topics:
 errors:
     on-failure: "skip"
 pipeline:
+  - name: "convert-to-structure"
+    input: "input-topic"
+    type: "document-to-json"
+    configuration:
+      text-field: "text"
   - name: "compute-embeddings"
     id: "step1"
     type: "compute-ai-embeddings"
-    input: "input-topic"
     output: "output-topic"
     configuration:
       model: "${secrets.open-ai.embeddings-model}" # This needs to match the name of the model deployment, not the base model
       embeddings-field: "value.embeddings"
-      text: "{{ value.name }} {{ value.description }}"
+      text: "{{ value.text }}"
       batch-size: 10
       # this is in milliseconds. It is important to take this value into consideration when using this agent in the chat response pipeline
       # in fact this value impacts the latency of the response