Skip to content

configuration.md

Latest commit

 

History

History
110 lines (81 loc) · 3.14 KB

configuration.md

File metadata and controls

110 lines (81 loc) · 3.14 KB

Configuration Guide

This guide covers the post-installation configuration steps needed to get Expanto fully operational with your data warehouse and AI services.

Legend:

  • ⭐ REQUIRED — must be configured for the system to work
  • 🔹 OPTIONAL — has sensible defaults, customize as needed

Overview

After running make setup, you'll have configuration templates in .streamlit/ directory that need to be customized for your environment:

  • .streamlit/secrets.toml — sensitive credentials (DWH, API keys)
  • .streamlit/expanto.toml — application settings
  • .streamlit/config.toml — Streamlit UI configuration

Required Configuration

1. Data Warehouse Connection ⭐ REQUIRED

Configure at least one data warehouse in .streamlit/secrets.toml:

BigQuery Setup

[bigquery]
file_path = "path/to/your-service-account.json"
project_name = "your-project-id"
connection_type = "service_account"  # or "application_default"

Steps:

  1. Create a service account in Google Cloud Console
  2. Download the JSON key file
  3. Grant BigQuery permissions: BigQuery Data Viewer, BigQuery Job User
  4. Update file_path with absolute path to JSON file
  5. Set connection_type = "service_account"

Snowflake Setup

[snowflake]
account = "your-account.region.cloud"
user = "your-username"
password = "your-password"
warehouse = "your-warehouse"
database = "your-database"
schema = "your-schema"

Steps:

  1. Ensure your Snowflake user has access to the target database/schema
  2. Test connection with the provided credentials
  3. Verify the account identifier format (include region and cloud provider)

2. AI Assistant Configuration ⭐ REQUIRED

API Keys ⭐ REQUIRED

Add your AI provider API key to .streamlit/secrets.toml:

[api_keys]
PROVIDER_API_KEY = "your-together-ai-api-key"  # ⭐ REQUIRED
TAVILY_API_KEY = "your-tavily-key"  # 🔹 OPTIONAL - for web search
LOGFIRE_TOKEN = "your-logfire-token"  # 🔹 OPTIONAL - for observability

Steps:

  1. ⭐ REQUIRED: Sign up at Together.ai and get an API key
  2. ⭐ REQUIRED: Add the key to PROVIDER_API_KEY
  3. 🔹 Optional: Get Tavily API key for web search capabilities
  4. 🔹 Optional: Get Logfire token for observability

Model Configuration 🔹 OPTIONAL

Customize AI models in .streamlit/expanto.toml (default models work fine):

[assistant.models]
fast = "deepseek-ai/DeepSeek-V3"
tool_thinker = "Qwen/Qwen3-235B-A22B-Thinking-2507"
agentic = "moonshotai/Kimi-K2-Instruct"AI Assistant Configuration

3. Database Selection ⭐ REQUIRED

Set which data warehouse to use for calculations in .streamlit/expanto.toml:

[precompute_db]
name = "snowflake"  # or "bigquery"

4. Internal Database 🔹 OPTIONAL

The SQLite configuration is pre-configured and works out of the box. Only customize if needed:

[internal_db]
engine_str = "sqlite:///expanto.db"
async_engine_str = "sqlite+aiosqlite:///expanto.db"

[internal_db.connect_args]
pool_size = 5
max_overflow = 5
pool_timeout = 10
pool_recycle = 900
connect_args = { check_same_thread = false }