MCP Integration

Connect Viete to over 8,000 data sources through MCP (Model Context Protocol), enabling automated data fetching for Excel model generation and updates. Pull data from databases, APIs, cloud platforms, SaaS tools, and virtually any digital system.

What is MCP?

Model Context Protocol (MCP) is an open standard that provides unified access to diverse data sources. Through MCP, Viete can connect to thousands of platforms without requiring custom integrations for each one.

How MCP Powers Viete Automation

Instead of manually exporting data from various systems and importing into Excel, Viete uses MCP to:

  • • Fetch data directly from source systems automatically
  • • Transform and structure data for Excel modeling
  • • Schedule regular data refreshes without manual intervention
  • • Maintain live connections to keep models current
  • • Combine data from multiple sources into unified models

8,000+ Supported Data Sources

MCP's extensive ecosystem means Viete can connect to virtually any platform you use:

Databases

  • • PostgreSQL, MySQL, SQL Server
  • • MongoDB, DynamoDB, Firestore
  • • Oracle, Snowflake, BigQuery
  • • Redis, Cassandra, Elasticsearch
  • • Any ODBC/JDBC compatible database

Cloud Platforms

  • • AWS (S3, RDS, DynamoDB, Redshift)
  • • Google Cloud (BigQuery, Cloud Storage)
  • • Azure (SQL Database, Blob Storage)
  • • Cloudflare (R2, D1, Workers KV)
  • • DigitalOcean, Heroku, Vercel

Business Applications

  • • Salesforce, HubSpot, Zoho CRM
  • • QuickBooks, Xero, NetSuite
  • • Stripe, PayPal, Square
  • • Shopify, WooCommerce, Magento
  • • Jira, Asana, Monday.com

Marketing & Analytics

  • • Google Analytics, Adobe Analytics
  • • Facebook/Instagram Ads
  • • Google Ads, LinkedIn Ads
  • • Mailchimp, SendGrid, Klaviyo
  • • Mixpanel, Amplitude, Segment

File Storage

  • • Google Drive, Dropbox, OneDrive
  • • Box, AWS S3, Azure Blob
  • • FTP/SFTP servers
  • • SharePoint, Confluence
  • • Local file systems

Communication & HR

  • • Slack, Microsoft Teams, Discord
  • • Gmail, Outlook, Office 365
  • • BambooHR, Workday, ADP
  • • Zendesk, Intercom, Freshdesk
  • • Calendar APIs (Google, Outlook)
Note
This is just a sample of commonly used platforms. MCP supports thousands more through its extensible architecture and community-contributed connectors.

How to Connect Data Sources

Connection Process

1. Select Data Source

Choose from the catalog of 8,000+ supported platforms. Search by name or browse by category (databases, SaaS, cloud, etc.).

2. Authenticate

Provide connection credentials: API keys, OAuth tokens, database connection strings, or username/password as required by the platform.

3. Configure Data Access

Specify what data to fetch: tables, queries, API endpoints, file paths, or custom data retrieval logic.

4. Map to Excel Model

Define how source data should populate your Excel model: which fields go to which columns, any transformations needed, and update behavior.

5. Test Connection

Verify the connection works and data flows correctly before activating automation or workflows.

MCP Server Types

Viete supports two types of MCP server configurations, each suited for different integration scenarios:

HTTP Servers

URL-based connections to web APIs and cloud services. Most common for SaaS platforms and cloud-based data sources.

Configuration:

  • • Server URL (required)
  • • Authorization token (optional)
  • • OAuth credentials (optional, encrypted)

Best For:

  • • REST APIs and webhooks
  • • Cloud-based services
  • • Third-party integrations
  • • OAuth-authenticated platforms

STDIO Servers

Command-based integrations that execute local programs or scripts. Perfect for databases, file systems, and custom data extraction tools.

Configuration:

  • • Command to execute (required)
  • • Command arguments (optional)
  • • Environment variables (optional)
  • • Secrets (encrypted)

Best For:

  • • Database connections
  • • Local file processing
  • • Custom scripts and tools
  • • Command-line utilities
Note
Choose the server type based on your data source. Most cloud services use HTTP, while databases and local systems typically use STDIO. Each server can be enabled or disabled independently.

Use Cases

Financial Dashboard Automation

Connect to accounting system (QuickBooks API), bank feeds (Plaid), and payment processors (Stripe API). Automatically generate daily P&L with real-time data from all sources.

Data Sources: QuickBooks, Stripe, Bank API via MCP

Frequency: Daily at 8 AM

Output: Comprehensive financial dashboard with actuals vs budget

Marketing Performance Tracking

Pull campaign data from Google Ads, Facebook Ads, LinkedIn Ads, and website analytics. Consolidate into unified marketing ROI dashboard updated hourly.

Data Sources: Google Ads API, Meta Ads API, LinkedIn API, Google Analytics via MCP

Frequency: Hourly

Output: Cross-platform campaign performance analysis

E-commerce Operations Dashboard

Connect to Shopify for sales data, inventory system for stock levels, shipping API for fulfillment metrics. Generate comprehensive operations dashboard.

Data Sources: Shopify API, inventory database, ShipStation API via MCP

Frequency: Every 4 hours

Output: Real-time operations dashboard with sales, inventory, fulfillment

Customer Success Metrics

Aggregate customer data from CRM (Salesforce), support tickets (Zendesk), product usage (internal database), and billing (Stripe). Track customer health scores and churn risk.

Data Sources: Salesforce API, Zendesk API, PostgreSQL, Stripe API via MCP

Frequency: Weekly

Output: Customer health dashboard with engagement metrics and renewal forecasts

Multi-Source Data Consolidation

One of MCP's most powerful capabilities is combining data from multiple sources into unified Excel models:

Example: Complete Business Overview

Revenue Data

From Stripe, PayPal, and Shopify via MCP

Expense Data

From QuickBooks and corporate credit card APIs via MCP

Customer Metrics

From Salesforce CRM via MCP

Marketing Data

From Google Analytics and ad platforms via MCP

Result

Comprehensive executive dashboard pulling from 7+ data sources, updated automatically, showing complete business health in one Excel model.

Tip
MCP handles data transformation and formatting automatically. Different source formats (JSON from APIs, CSV from exports, database rows) are all normalized into clean Excel-ready data.

Security & Credentials

Viete prioritizes security for all MCP connections with comprehensive encryption and access controls:

Encryption at Every Layer

  • • Authorization tokens encrypted before database storage
  • • OAuth client IDs and secrets encrypted at rest
  • • STDIO server secrets and environment variables encrypted
  • • All credentials transmitted over encrypted connections only

OAuth Support

For platforms supporting OAuth authentication (like GitHub, Zapier, and many SaaS tools), Viete provides built-in OAuth flow:

  • • Standard OAuth 2.0 authorization flow
  • • Encrypted storage of OAuth tokens
  • • Automatic token refresh handling
  • • Secure redirect handling

Access Control

  • • MCP servers scoped to individual users
  • • Enable/disable servers independently
  • • Revoke access anytime from settings
  • • Credentials never exposed in Excel outputs or logs
Note
Best Practice: Always use read-only credentials or API keys when possible. Viete only needs to fetch data, not modify it in source systems. This follows the principle of least privilege for maximum security.

Data Refresh Frequency

For automation, you can set how often Viete should fetch fresh data via MCP:

Real-Time

Every 15 minutes to 1 hour for time-sensitive dashboards and monitoring.

Daily

Once or multiple times per day for operational reports and standard dashboards.

Periodic

Weekly, monthly, or custom schedules for less time-sensitive analysis.

Custom APIs & Internal Systems

Beyond the 8,000+ pre-built connectors, MCP can connect to custom APIs and internal systems:

  • REST APIs with custom authentication
  • GraphQL endpoints
  • SOAP web services
  • Internal databases on private networks
  • Custom data pipelines and ETL outputs
  • Proprietary systems with API access

If your system has an API or database interface, MCP can likely connect to it— even if it's not in the standard catalog.

Best Practices

  • Test connections with small data samples before full automation
  • Use read-only credentials whenever possible
  • Document which data sources feed which Excel models
  • Set appropriate refresh frequencies (avoid over-fetching)
  • Monitor API rate limits for high-frequency automations
  • Keep credentials updated when API keys rotate

Related