Using the IceCubes API and MCP Server to Build Custom Meeting Data Integrations

IceCubes includes built-in integrations for HubSpot, Salesforce, Slack, and Zapier. For many teams, these cover the primary workflows. But some organizations need meeting data flowing into places those integrations do not reach: internal BI platforms, custom dashboards, data warehouses, proprietary CRM systems, or AI tools that need meeting context.

That is what the IceCubes API and MCP server are for. This guide covers practical integration patterns for teams that want to build custom connections between their meeting data and their internal tools.

For the full API and MCP reference, see our API and MCP Server overview.

When to Build a Custom Integration

Before building something custom, check whether an existing integration or Zapier workflow covers your use case. Custom integrations make sense when:

• You need meeting data in a system that IceCubes does not have a native integration for (e.g., a proprietary CRM, an internal wiki, or a BI tool like Looker or Metabase)
• You want to combine meeting data with other data sources for custom analytics
• You need to trigger internal workflows based on specific meeting events or Smart Tag results
• You are building an AI-powered tool that needs meeting context via MCP
• You want to feed meeting data into a data warehouse for long-term analysis

Integration Pattern 1: Meeting Data to BI Platforms

Use case: Your revenue operations team wants to correlate meeting activity with pipeline outcomes. How many meetings does it take to close a deal? Do deals with more discovery questions close at higher rates? Which reps have the highest meeting-to-close conversion?

How to build it:

1. Set up a scheduled job (daily or weekly) that calls the IceCubes API to pull recent meetings
2. For each meeting, extract the metadata you care about: date, duration, participants, platform, Smart Tag results, MEDDIC completion, action item count
3. Load the data into your BI platform (Looker, Metabase, Tableau, Power BI) or data warehouse (Snowflake, BigQuery, Redshift)
4. Join meeting data with CRM pipeline data to build your analytics

Key API endpoints:

• GET /api/v1/meetings with date range filters to pull recent meetings
• GET /api/v1/meetings/:id/insights for MEDDIC scores, Smart Tags, and action items
• GET /api/v1/insights/smart-tags for aggregate Smart Tag data across meetings

Example metrics to track:

Integration Pattern 2: Custom Internal Dashboards

Use case: Your sales manager wants a real-time dashboard showing today's meeting activity, open action items, and Smart Tag alerts, without logging into IceCubes separately.

How to build it:

Build a lightweight dashboard (Retool, Streamlit, or your own frontend) that calls the IceCubes API on page load:

1. GET /api/v1/meetings?date_from=today to show today's meetings and their summaries
2. GET /api/v1/insights/action-items?status=pending to display all open action items with assignees
3. GET /api/v1/insights/smart-tags?date_from=this_week to show this week's Smart Tag activity

Dashboard sections to consider:

• Activity feed: Recent meetings with one-line summaries
• Action items: Open items grouped by assignee, sorted by due date
• Signal alerts: Smart Tag results from the current week, highlighting churn risk, expansion signals, and competitor mentions
• Team metrics: Meeting counts, average duration, and AI credit usage per team member

Integration Pattern 3: Automated Workflows from Meeting Events

Use case: When a customer meeting triggers a "Churn Risk" Smart Tag, automatically create a task in your project management tool, notify the account manager's supervisor, and add the account to a watch list.

How to build it:

1. Configure a webhook listener for the meeting.completed event from IceCubes
2. When a new meeting arrives, check the Smart Tag results via the API
3. If specific tags are present, trigger your workflow:
   • Create a task in Jira, Asana, or Linear
   • Send a notification via your internal messaging system
   • Update a status field in your CRM or internal database
   • Add the account to a review queue for the next team meeting

This goes beyond what Zapier workflows can do because you can implement custom logic: only trigger if the churn risk tag fired more than twice in the last 30 days, or only if the account's ARR exceeds a threshold, or only if the customer's renewal date is within 90 days.

Integration Pattern 4: AI Tools via MCP

Use case: Your team uses Claude, ChatGPT, or a custom LLM application for daily work. You want these AI tools to have access to your meeting data so they can answer questions like "What did the client say about our pricing?" or "Summarize all customer feedback from this week's calls."

How MCP works:

The MCP (Model Context Protocol) server exposes your IceCubes meeting data as tools that AI assistants can call. When you ask Claude "What action items are assigned to me this week?", Claude calls the IceCubes MCP server's get_action_items tool, retrieves the data, and responds with a formatted answer.

Available MCP tools:

• search_meetings for finding meetings by keyword, date, participant, or tag
• get_meeting for full meeting details including transcript
• get_insights for AI-generated insights on a specific meeting
• get_action_items for listing action items with filters
• get_smart_tags for querying Smart Tag results

Practical MCP use cases:

• Meeting prep: "What were the key points from our last three meetings with Acme Corp?"
• Action item review: "List all action items assigned to me that are overdue"
• Competitive intelligence: "Which customers have mentioned [competitor] in the last month?"
• Account health check: "Summarize the sentiment across our recent meetings with [customer]"
• Custom reporting: "Generate a weekly summary of all Smart Tag activity across my team"

Integration Pattern 5: Data Warehouse for Long-Term Analysis

Use case: Your data team wants meeting data in your data warehouse alongside CRM data, product usage data, and support ticket data for comprehensive customer health analysis.

How to build it:

1. Set up a nightly ETL job that pulls all meetings from the last 24 hours via the API
2. Normalize the data into tables: meetings, participants, action_items, smart_tags, insights
3. Load into your data warehouse (Snowflake, BigQuery, Redshift)
4. Build models that join meeting data with other customer data sources

Analysis you can run with meeting data in your warehouse:

• Correlate meeting frequency and sentiment with customer retention rates
• Identify which types of meetings (QBRs, demos, training) are most associated with expansion revenue
• Track which Smart Tag signals are most predictive of churn
• Measure rep effectiveness based on meeting outcomes across their entire book of business

Best Practices for Custom Integrations

Respect rate limits. Check the API docs for current rate limits. Use pagination for large result sets rather than pulling everything at once.

Cache when possible. If your dashboard refreshes every 5 minutes, cache meeting data locally rather than hitting the API on every page load. Meeting data does not change after it is generated, so caching is safe.

Use webhooks for real-time workflows. If you need to react to meetings as they happen, use webhook events rather than polling the API on a schedule.

Start with one integration. Do not try to connect meeting data to five systems at once. Pick the highest-impact integration, build it, prove the value, and then expand.

Monitor and alert. Set up monitoring for your integration. If the ETL job fails or the webhook stops receiving events, you want to know immediately, not when someone notices the dashboard is stale.

Getting Started

Generate your API key in Settings > Integrations > API. Full API documentation is available at /docs/api. If you are not yet using IceCubes, start with 50 free AI credits, no credit card required.

Explore the API Docs | Add to Chrome | Add to Edge