Feature product telemetry (#10)

Author: nithinjamespallicka

.gitignore

@@ -61,4 +61,5 @@ datu-sqlserver/
 test_schema_cache.json
 
 .cache/
-site/
+site/
+.coverage*

Dockerfile

@@ -9,8 +9,8 @@ RUN apt-get update && \
     rm -rf /var/lib/apt/lists/* && \
     pip install --upgrade pip
 
-RUN curl https://packages.microsoft.com/keys/microsoft.asc | apt-key add -
-RUN curl https://packages.microsoft.com/config/debian/11/prod.list > /etc/apt/sources.list.d/mssql-release.list
+RUN curl https://packages.microsoft.com/keys/microsoft.asc | gpg --dearmor > /etc/apt/trusted.gpg.d/microsoft.gpg \
+&& curl https://packages.microsoft.com/config/debian/11/prod.list -o /etc/apt/sources.list.d/mssql-release.list
 
 RUN apt-get update
 RUN env ACCEPT_EULA=Y apt-get install -y msodbcsql18
@@ -21,4 +21,4 @@ RUN uv sync --extra postgres --extra sqldb
 ENV PATH="/app/.venv/bin:$PATH"
 # Reset the entrypoint, don't invoke `uv`
 ENTRYPOINT [] 
-CMD ["uvicorn", "datu.main:app", "--host", "0.0.0.0", "--port", "8000", "--reload"]
+CMD ["uvicorn", "datu.main:app", "--host", "0.0.0.0", "--port", "8000", "--reload"]

changelog.d/+7413cd44.added.md

@@ -0,0 +1 @@
+Enable product Telemetry

pyproject.toml

@@ -38,6 +38,7 @@ dependencies = [
     "fastmcp>=2.10.5",
     "mcp-use[search]>=1.3.7",
     "onnxruntime==1.19.2 ; sys_platform == 'darwin' and platform_machine == 'x86_64'",
+    "posthog>=6.5.0",
 ]
 
 [project.urls]

src/datu/app_config.py

@@ -17,6 +17,7 @@
 from datu.integrations.config import IntegrationConfigs
 from datu.mcp.config import MCPConfig
 from datu.services.config import SchemaRAGConfig
+from datu.telemetry.config import TelemetryConfig
 
 
 class Environment(Enum):
@@ -56,6 +57,13 @@ class DatuConfig(BaseSettings):
         schema_sample_limit (int): The maximum number of rows to sample from the schema.
         schema_categorical_threshold (int): The threshold for categorical columns in the schema.
         enable_schema_rag (bool): Enable RAG for schema extraction.
+        enable_anonymized_telemetry (bool): Enable anonymized telemetry. Default is True.
+        app_environment (str): The application environment (e.g., "dev", "test", "prod"). Default is "dev".
+        telemetry (TelemetryConfig | None): Configuration settings for telemetry.
+        enable_mcp (bool): Whether to enable MCP integration. Default is False.
+        mcp (MCPConfig | None): Configuration settings for MCP integration.
+        enable_schema_rag (bool): Enable RAG for schema extraction.
+        schema_rag (SchemaRAGConfig | None): Configuration settings for schema RAG.
 
     Attributes:
         host (str): The host address for the application.
@@ -77,7 +85,8 @@ class DatuConfig(BaseSettings):
         mcp (MCPConfig | None): Configuration settings for MCP integration.
         enable_schema_rag (bool): Enable RAG for schema extraction.
         schema_rag (SchemaRAGConfig | None): Configuration settings for schema RAG.
-
+        enable_anonymized_telemetry (bool): Enable anonymized telemetry.
+        telemetry (TelemetryConfig | None): Configuration settings for telemetry.
 
     """
 
@@ -110,7 +119,8 @@ class DatuConfig(BaseSettings):
         description="Configuration settings for schema RAG (Retrieval-Augmented Generation).",
     )
     enable_anonymization: bool = False
-
+    enable_anonymized_telemetry: bool = True
+    telemetry: TelemetryConfig | None = Field(default_factory=TelemetryConfig)
     model_config = SettingsConfigDict(
         env_prefix="datu_",
         env_nested_delimiter="__",

src/datu/factory/llm_client_factory.py

@@ -6,6 +6,8 @@
 from typing import Literal
 
 from datu.llm_clients.openai_client import OpenAIClient
+from datu.telemetry.product.events import MCPClientEvent
+from datu.telemetry.product.posthog import get_posthog_client
 
 
 def get_llm_client(provider: Literal["openai"] | None = None) -> OpenAIClient | None:
@@ -20,6 +22,15 @@ def get_llm_client(provider: Literal["openai"] | None = None) -> OpenAIClient |
         ValueError: If the specified provider is not supported.
     """
     if provider == "openai":
-        return OpenAIClient()
+        openai_client = OpenAIClient()
+        if openai_client.agent:
+            posthog_client = get_posthog_client()
+            if openai_client.mcp_client and getattr(openai_client.mcp_client, "config", None):
+                servers = openai_client.mcp_client.config.get("mcpServers", {})
+                server_names = list(servers.keys()) if servers else []
+            else:
+                server_names = []
+            posthog_client.capture(MCPClientEvent(server_names=server_names))
+        return openai_client
     else:
         raise ValueError("Invalid LLM provider specified in configuration.")

src/datu/main.py

@@ -13,6 +13,8 @@
 from datu.app_config import get_logger, settings
 from datu.routers import chat, metadata, transformations
 from datu.schema_extractor.schema_cache import load_schema_cache
+from datu.telemetry.product.events import OpenAIEvent
+from datu.telemetry.product.posthog import get_posthog_client
 
 logger = get_logger(__name__)
 
@@ -56,6 +58,8 @@ def start_app() -> None:
     It also sets the logging level based on the configuration settings.
     """
     logger.info("Starting the FastAPI application...")
+    posthog_client = get_posthog_client()
+    posthog_client.capture(OpenAIEvent())
     uvicorn.run(app, host=settings.host, port=settings.port)
 
 

tests/helpers/__init__.py src/datu/telemetry/__init__.pytests/helpers/__init__.py renamed to src/datu/telemetry/__init__.py

@@ -0,0 +1,19 @@
+"""Telemetry configuration settings."""
+
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class TelemetryConfig(BaseSettings):
+    """Telemetry configuration settings."""
+
+    api_key: str = "phc_m74dfR9nLpm2nipvkL2swyFDtNuQNC9o2FL2CSbh6Je"
+    package_name: str = "datu-core"
+
+    model_config = SettingsConfigDict(
+        env_prefix="telemetry_",
+        env_nested_delimiter="__",
+    )
+
+
+def get_telemetry_settings() -> TelemetryConfig:
+    return TelemetryConfig()