Connect to Every Tool Your Team Uses
OperativeOps agents pull real-time data from the tools you already use. Native integrations, MCP protocol, and a custom SDK — connect anything.
Native Integrations
Pre-Built Connectors
Connect in minutes. Each integration is optimized for the data types your AI agents need.
Slack
Surface agent insights directly in Slack channels. Agents respond to @mentions and post proactive alerts.
Agent use: All agents can post updates, answer questions, and trigger workflows.
Jira
Track engineering velocity, sprint progress, bug metrics, and tech debt scores in real time.
Agent use: Alex (CTO) uses Jira data for technical assessments and sprint analysis.
Salesforce
Revenue metrics, pipeline data, deal velocity, and win rate analysis for business intelligence.
Agent use: Jordan (CEO) and Riley (Analytics) use pipeline data for revenue forecasting.
Notion
Sync company knowledge base, OKRs, meeting notes, and strategic documentation.
Agent use: All agents reference Notion for institutional knowledge and strategic context.
Google Drive
Index documents, spreadsheets, and presentations across your organization.
Agent use: RAG engine indexes Drive content for cross-agent knowledge retrieval.
HubSpot
Marketing performance, lead generation, campaign analytics, and attribution data.
Agent use: Sam (Marketing) uses HubSpot data for campaign ROI and growth insights.
Microsoft Teams
Bring OperativeOps agents into your Teams workspace for inline intelligence.
Agent use: All agents available via Teams bot with same capabilities as Slack integration.
PostgreSQL
Connect directly to your database for real-time metrics, custom KPIs, and data analysis.
Agent use: Riley (Analytics) queries your data directly for custom analysis and forecasting.
Custom Integrations
Build Your Own with MCP SDK
The Model Context Protocol SDK lets you connect any data source or tool to OperativeOps. Define tools, resources, and prompts that agents can use — in TypeScript, Python, or Go.
- Define custom tools agents can invoke
- Expose data as MCP resources for RAG indexing
- Create custom prompt templates
- Full TypeScript, Python, and Go SDK support
- Webhook-based or polling data sync
import { McpServer } from "@operativeops/mcp-sdk";
const server = new McpServer({
name: "internal-wiki",
version: "1.0.0",
});
server.addTool({
name: "search_wiki",
description: "Search internal wiki",
parameters: { query: "string" },
async handler({ query }) {
return searchWiki(query);
},
});
server.start();Data Modes
Sync vs. Live Mode
Choose how your integrations deliver data to agents — batch syncing for efficiency or live connections for real-time insights.
Sync Mode
Scheduled batch updates
Data is synced on a schedule (every 5 min to every 24 hrs). Best for large datasets, historical analysis, and lower API usage. Agents work with the most recent snapshot.
- Configurable sync intervals
- Full historical data indexing
- Lower API quota usage
- Best for: Salesforce, HubSpot, PostgreSQL
Live Mode
Real-time event streaming
Data streams in real time via webhooks or event subscriptions. Best for chat integrations, alerts, and time-sensitive workflows where agents need instant context.
- Sub-second data freshness
- Webhook and event-driven
- Streaming agent responses
- Best for: Slack, Teams, Jira webhooks