[agents] Add support for AstraDB Collections (Astra Vector DB, using …

…Stargate) (#731)

examples/applications/chatbot-rag-memory/crawler.yaml

@@ -46,7 +46,6 @@ pipeline:
   - name: "Detect language"
     type: "language-detector"
     configuration:
-      allowedLanguages: ["en"]
       property: "language"
   - name: "Split into chunks"
     type: "text-splitter"

examples/applications/flare/crawler.yaml

@@ -70,7 +70,6 @@ pipeline:
   - name: "Detect language"
     type: "language-detector"
     configuration:
-      allowedLanguages: ["en", "fr"]
       property: "language"
   - name: "Split into chunks"
     type: "text-splitter"

examples/applications/langchain-chat/crawler.yaml

@@ -50,7 +50,6 @@ pipeline:
   - name: "Detect language"
     type: "language-detector"
     configuration:
-      allowedLanguages: ["en", "fr"]
       property: "language"
   - name: "Split into chunks"
     type: "text-splitter"

examples/applications/query-milvus/crawler.yaml

@@ -106,7 +106,6 @@ pipeline:
   - name: "Detect language"
     type: "language-detector"
     configuration:
-      allowedLanguages: ["en", "fr"]
       property: "language"
   - name: "Split into chunks"
     type: "text-splitter"

examples/applications/query-solr/crawler.yaml

@@ -102,7 +102,6 @@ pipeline:
   - name: "Detect language"
     type: "language-detector"
     configuration:
-      allowedLanguages: ["en", "fr"]
       property: "language"
   - name: "Split into chunks"
     type: "text-splitter"

examples/applications/webcrawler-astra-vector-db/.gitignore

@@ -0,0 +1 @@
+java/lib/*

examples/applications/webcrawler-astra-vector-db/README.md

@@ -0,0 +1,45 @@
+# Indexing a WebSite
+
+This sample application shows how to use the WebCrawler Source Connector and Astra Vector DB.
+
+## Collections
+
+This application creates a collection named "documents" in your DB.
+You can change the name of the collection in the file `configuration.yaml`.
+
+## Configure access to the Vector Database
+
+Export some ENV variables in order to configure access to the database:
+
+```
+export ASTRA_VECTOR_DB_TOKEN=AstraCS:...
+export ASTRA_VECTOR_DB_ENDPOINT=https://....astra.datastax.com
+```
+
+You can find the credentials in the Astra DB console.
+
+The examples/secrets/secrets.yaml resolves those environment variables for you.
+When you go in production you are supposed to create a dedicated secrets.yaml file for each environment.
+
+## Configure the pipeline
+
+You can edit the file `crawler.yaml` and configure the list of the allowed web domains, this is required in order to not let the crawler escape outside your data.
+Configure the list of seed URLs, for instance with your home page.
+
+The default configuration in this example will crawl the LangStream website.
+
+## Deploy the LangStream application
+
+```
+./bin/langstream docker run test -app examples/applications/webcrawler-astra-vector-db  -s examples/secrets/secrets.yaml
+```
+
+## Talk with the Chat bot
+
+If the application launches successfully, you can talk with the chat bot using UI.
+
+You can also use the CLI:
+
+```
+./bin/langstream gateway chat test -cg bot-output -pg user-input -p sessionId=$(uuidgen)
+```

examples/applications/webcrawler-astra-vector-db/chatbot.yaml

@@ -0,0 +1,106 @@
+#
+# Copyright DataStax, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+topics:
+  - name: "questions-topic"
+    creation-mode: create-if-not-exists
+  - name: "answers-topic"
+    creation-mode: create-if-not-exists
+  - name: "log-topic"
+    creation-mode: create-if-not-exists
+errors:
+    on-failure: "skip"
+pipeline:
+  - name: "convert-to-structure"
+    type: "document-to-json"
+    input: "questions-topic"
+    configuration:
+      text-field: "question"
+  - name: "compute-embeddings"
+    type: "compute-ai-embeddings"
+    configuration:
+      model: "${secrets.open-ai.embeddings-model}" # This needs to match the name of the model deployment, not the base model
+      embeddings-field: "value.question_embeddings"
+      text: "{{ value.question }}"
+      flush-interval: 0
+  - name: "lookup-related-documents"
+    type: "query-vector-db"
+    configuration:
+      datasource: "AstraDatasource"
+      query: |
+          {
+              "collection-name": "${globals.collection-name}",
+              "limit": 20,
+              "vector": ?
+          }
+      fields:
+        - "value.question_embeddings"
+      output-field: "value.related_documents"
+  - name: "re-rank documents with MMR"
+    type: "re-rank"
+    configuration:
+      max: 5 # keep only the top 5 documents, because we have an hard limit on the prompt size
+      field: "value.related_documents"
+      query-text: "value.question"
+      query-embeddings: "value.question_embeddings"
+      output-field: "value.related_documents"
+      text-field: "record.text"
+      embeddings-field: "record.vector"
+      algorithm: "MMR"
+      lambda: 0.5
+      k1: 1.2
+      b: 0.75
+  - name: "ai-chat-completions"
+    type: "ai-chat-completions"
+    configuration:
+      model: "${secrets.open-ai.chat-completions-model}" # This needs to be set to the model deployment name, not the base name
+      # on the log-topic we add a field with the answer
+      completion-field: "value.answer"
+      # we are also logging the prompt we sent to the LLM
+      log-field: "value.prompt"
+      # here we configure the streaming behavior
+      # as soon as the LLM answers with a chunk we send it to the answers-topic
+      stream-to-topic: "answers-topic"
+      # on the streaming answer we send the answer as whole message
+      # the 'value' syntax is used to refer to the whole value of the message
+      stream-response-completion-field: "value"
+      # we want to stream the answer as soon as we have 20 chunks
+      # in order to reduce latency for the first message the agent sends the first message
+      # with 1 chunk, then with 2 chunks....up to the min-chunks-per-message value
+      # eventually we want to send bigger messages to reduce the overhead of each message on the topic
+      min-chunks-per-message: 20
+      messages:
+        - role: system
+          content: |
+              An user is going to perform a questions, The documents below may help you in answering to their questions.
+              Please try to leverage them in your answer as much as possible.
+              Take into consideration that the user is always asking questions about the LangStream project.
+              If you provide code or YAML snippets, please explicitly state that they are examples.
+              Do not provide information that is not related to the LangStream project.
+            
+              Documents:
+              {{# value.related_documents}}
+              {{ text}}
+              {{/ value.related_documents}}
+        - role: user
+          content: "{{ value.question}}"
+  - name: "cleanup-response"
+    type: "drop-fields"
+    output: "log-topic"
+    configuration:
+      fields:
+        - "question_embeddings"
+        - "related_documents"

examples/applications/webcrawler-astra-vector-db/configuration.yaml

@@ -0,0 +1,34 @@
+#
+#
+# Copyright DataStax, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+configuration:
+  defaults:
+    globals:
+      collection-name: "documents"
+  resources:
+  - type: "open-ai-configuration"
+    name: "OpenAI Azure configuration"
+    configuration:
+      url: "${secrets.open-ai.url}"
+      access-key: "${secrets.open-ai.access-key}"
+      provider: "${secrets.open-ai.provider}"
+  - type: "datasource"
+    name: "AstraDatasource"
+    configuration:
+      service: "astra-vector-db"
+      token: "${secrets.astra-vector-db.token}"
+      endpoint: "${secrets.astra-vector-db.endpoint}"

examples/applications/webcrawler-astra-vector-db/crawler.yaml

@@ -0,0 +1,93 @@
+#
+# Copyright DataStax, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+name: "Crawl a website"
+topics:
+  - name: "chunks-topic"
+    creation-mode: create-if-not-exists
+resources:
+      size: 2
+pipeline:
+  - name: "Crawl the WebSite"
+    type: "webcrawler-source"
+    configuration:
+      seed-urls: ["https://docs.langstream.ai/"]
+      allowed-domains: ["https://docs.langstream.ai"]
+      forbidden-paths: []
+      min-time-between-requests: 500
+      reindex-interval-seconds: 3600
+      max-error-count: 5
+      max-urls: 1000
+      max-depth: 50
+      handle-robots-file: true
+      user-agent: "" # this is computed automatically, but you can override it
+      scan-html-documents: true
+      http-timeout: 10000
+      handle-cookies: true
+      max-unflushed-pages: 100
+      # store data directly on the agent disk, no need for external S3 storage
+      state-storage: disk
+  - name: "Extract text"
+    type: "text-extractor"
+  - name: "Normalise text"
+    type: "text-normaliser"
+    configuration:
+      make-lowercase: true
+      trim-spaces: true
+  - name: "Detect language"
+    type: "language-detector"
+    configuration:
+      property: "language"
+  - name: "Split into chunks"
+    type: "text-splitter"
+    configuration:
+      splitter_type: "RecursiveCharacterTextSplitter"
+      chunk_size: 400
+      separators: ["\n\n", "\n", " ", ""]
+      keep_separator: false
+      chunk_overlap: 100
+      length_function: "cl100k_base"
+  - name: "Convert to structured data"
+    type: "document-to-json"
+    configuration:
+      text-field: text
+      copy-properties: true
+  - name: "prepare-structure"
+    type: "compute"
+    configuration:
+      fields:
+        - name: "value.filename"
+          expression: "properties.url"
+          type: STRING
+        - name: "value.chunk_id"
+          expression: "properties.chunk_id"
+          type: STRING
+        - name: "value.language"
+          expression: "properties.language"
+          type: STRING
+        - name: "value.chunk_num_tokens"
+          expression: "properties.chunk_num_tokens"
+          type: STRING
+  - name: "compute-embeddings"
+    id: "step1"
+    type: "compute-ai-embeddings"
+    output: "chunks-topic"
+    configuration:
+      model: "${secrets.open-ai.embeddings-model}"
+      embeddings-field: "value.embeddings_vector"
+      text: "{{ value.text }}"
+      batch-size: 10
+      flush-interval: 500

examples/applications/webcrawler-astra-vector-db/gateways.yaml

@@ -0,0 +1,43 @@
+#
+#
+# Copyright DataStax, Inc.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+gateways:
+  - id: "user-input"
+    type: produce
+    topic: "questions-topic"
+    parameters:
+      - sessionId
+    produceOptions:
+      headers:
+        - key: langstream-client-session-id
+          valueFromParameters: sessionId
+
+  - id: "bot-output"
+    type: consume
+    topic: "answers-topic"
+    parameters:
+      - sessionId
+    consumeOptions:
+      filters:
+        headers:
+          - key: langstream-client-session-id
+            valueFromParameters: sessionId
+
+
+  - id: "llm-debug"
+    type: consume
+    topic: "log-topic"