Setup
Welcome to Teela! This guide walks you through the entire setup process, from connecting your first data source to inviting your team. By the end, Teela will be ready to answer questions about your data in plain English.
The Setup Process at a Glance
Getting Teela ready for your organization follows five steps:
- Connect your data: Add a database or file-based connection so Teela has something to work with.
- Extract the schema: Teela analyzes your database to learn about tables, columns, and relationships.
- Train the vocabulary: Walk through the Vocabulary Wizard so Teela understands your business terminology.
- Validate: Ask test questions to confirm Teela generates correct queries.
- Invite users: Bring your team on board and assign them to the right connections.
You don't have to complete everything in one sitting. Teela saves your progress, and you can return to any step at any time from the Admin portal.
Step 1: Connect Your Data
Database Connections (SQL Server, MySQL, PostgreSQL)
To connect a database, you'll need the following information from your IT team or database administrator:
- Hostname: The server address (e.g.,
db.yourcompany.comor an IP address) - Port: The port number (defaults: SQL Server 1433, MySQL 3306, PostgreSQL 5432)
- Database name: The specific database Teela should connect to
- Username and password: We strongly recommend using a read-only database account. Teela only needs to read your data, never write to it.
Adding the Connection
- In the Admin portal, navigate to Team Connections and click Add Connection.
- Select your database type (SQL Server, MySQL, or PostgreSQL).
- Enter your hostname, port, database name, and credentials.
- Click Test Connection to verify everything works before saving. Teela will attempt to connect and confirm it can reach your database.
- Once the test passes, click Save.
Your credentials are encrypted at rest and are never displayed in the admin interface after saving. No one can view them once saved.
Get Help from IT
If you don't have database credentials yourself, Teela makes it easy to involve your IT team without giving them a Teela account:
- On the connection setup screen, click Get Help from IT.
- Enter your IT contact's email address.
- They'll receive a secure link where they can enter the database credentials directly.
- The link expires after 72 hours for security.
- Once they submit the credentials, you'll be notified and the connection will be ready to configure.
This flow keeps credentials secure: they go directly into Teela's encrypted storage without passing through email or chat.
File-Based Connections (Google Sheets, Excel, CSV)
You can also connect Teela to spreadsheet data for your entire organization. For details on each file type, see Managing Connections.
- Google Sheets: Connect via OAuth, select the sheet and tabs you want, and Teela will auto-sync when data changes.
- Excel: Upload an
.xlsxfile (up to 100 MB) and select which sheets to include. - CSV: Upload a
.csvfile (up to 100 MB).
Step 2: Extract the Schema
Once your connection is saved, it's time for Teela to learn about your data structure.
- Navigate to your connection in the Admin portal and click Extract Schema.
- Teela will analyze your database to identify tables, columns, data types, relationships, and data patterns.
- For large databases, extraction runs in the background. You'll receive an email notification when it's complete.
After extraction finishes, you can review everything Teela discovered in the Data Dictionary and the interactive Schema Diagram.
Schema extraction is also a good time to exclude tables Teela doesn't need: like staging tables, logs, or system tables. You can do this from the Schema Diagram. See Training the AI for details.
Step 3: Train the Vocabulary with the Vocabulary Wizard
After schema extraction, Teela launches the Vocabulary Wizard: a guided walkthrough that helps Teela understand your business terminology.
Here's how it works:
- Teela generates a series of AI-powered questions organized around common business pillars like People, Customers, Products, and Finances.
- You answer each question in plain English. For example, Teela might ask: "What do you call your customers in everyday conversation?" and you'd answer something like "We call them clients."
- As you answer, Teela automatically creates aliases that map your business terms to the correct database tables and columns.
- At the end of the wizard, you'll get a Test Drive: a chance to ask a few sample questions and see how Teela responds.
The Vocabulary Wizard is optional but highly recommended. It dramatically improves Teela's accuracy right out of the gate.
Step 4: Validate Your Training
Before inviting your team, take a few minutes to test Teela's understanding.
- Open the Training Validation chat in the Admin portal.
- Ask questions your team would typically ask: things like "How many active customers do we have?" or "What was total revenue last quarter?"
- Review the SQL queries Teela generates. Are they hitting the right tables and columns?
- If something looks off, you can add documentation, SQL examples, or aliases to improve accuracy. See Training the AI for the full training toolkit.
A few minutes of validation here saves a lot of confusion later. Aim for at least 5-10 test questions covering your most common reporting needs.
Step 5: Invite Users
Once you're satisfied with Teela's accuracy, it's time to bring your team on board.
- Go to Users in the Admin portal and click Invite User.
- Enter their email address and select a role (typically User for most team members).
- Assign them to one or more connections: users can only see and query the connections you assign to them.
- They'll receive an email invitation with a link to get started.
For more details on roles, permissions, and security settings, see User Management.
What's Next?
Your initial setup is complete, but Teela gets smarter over time. Here are the resources you'll want to explore as you continue:
- Managing Connections: Add more data sources, manage sync settings, and handle file replacements.
- Connect Slack: Let your team ask Teela questions directly from Slack channels and DMs.
- Training the AI: Deepen Teela's understanding with documentation, SQL examples, and aliases.
- Monitoring and Improving: Review learning candidates, track knowledge gaps, and validate training changes.
- User Management: Manage roles, enforce MFA, and configure security policies.
- Aliases: Map business terms to database columns so your team can ask questions naturally.
- Data Dictionary: Review and enhance the descriptions Teela uses to understand your schema.