Insighto.ai

@insighto.ai show me all conversations from the last 7 days that are still unresolved, grouped by contact priority

Based on the conversations retrieved from the past week, here's what I found: 23 unresolved conversations total. High-priority contacts (8 conversations) include issues around billing integration failures and onboarding blockers. Medium-priority contacts (11 conversations) mostly involve feature requests and clarification questions. Low-priority contacts (4 conversations) are general inquiries. The oldest unresolved thread is from 6 days ago with a high-priority enterprise customer regarding API rate limits.

This example demonstrates the MCP's conversation retrieval and filtering capabilities. Requires API key with read access to conversations. The grouping and priority analysis happens through the AI's reasoning layer — Insighto.ai returns the conversation list, and the AI synthesizes patterns. Be aware that large date ranges may hit pagination limits requiring multiple calls.

@insighto.ai create a new custom intent called 'request_demo' that triggers when users ask to see the product in action, and add training phrases like 'show me how it works' and 'can I see a demo'

I've created the 'request_demo' intent in your Insighto.ai workspace with the following configuration: Intent name set to 'request_demo', added 2 training phrases ('show me how it works', 'can I see a demo'), and marked it as active. The intent is now ready for additional training phrases or deployment to your assistant. You'll need to configure the response action separately — this intent creation establishes the trigger pattern but doesn't define what happens when it fires.

This showcases the MCP's intent management for conversational AI customization. Requires API key with write permissions. Creating an intent is just the first step — you'll still need to wire up the response flow (webhook, canned response, or routing rule) through Insighto.ai's interface. The MCP handles the intent structure but not the full conversational logic.

@insighto.ai pull the contact record for ID c_8x4j2p and check if they have any custom fields related to subscription tier or renewal date

Retrieved contact record for c_8x4j2p (Sarah Chen, sarah.chen@techcorp.example). Checking custom fields: Found 'subscription_tier' set to 'Enterprise', 'renewal_date' showing '2025-09-15', and 'account_health_score' at 87/100. Also present: 'primary_use_case' (Customer Support Automation) and 'onboarding_completed' (true). This contact appears to be an established enterprise customer with renewal coming up in about 5 months.

This example highlights both contact retrieval and custom field inspection — useful for enriching customer context before outreach or support. Requires a valid contact ID and read permissions. Custom field schemas vary by organization, so the fields shown here are illustrative. If you're building automation around custom fields, use the Read Contact Custom Field List tool first to discover your org's actual field definitions.

When Insighto wins for pre-launch chatbot training

A 6-person SaaS team launching in 8 weeks needs a customer-facing chatbot that answers product questions without burning support hours. Insighto's intent-creation and datasource tools let you wire up a widget that pulls from your docs and FAQ before you have real ticket volume. The trade-off: you're building custom intents by hand, which works when you have 20-30 question patterns but gets tedious past 50. If your product has fewer than 40 core help topics and you want the bot live at launch, Insighto gives you the scaffolding without waiting for ML training data. You'll spend 2-3 days on intent design, then the widget handles tier-1 questions from day one.

When contact-field mapping justifies the API overhead

A 12-person sales team uses a homegrown CRM and needs to push conversation metadata back into contact records after demos. Insighto's contact and custom-field tools let you read the schema, match fields, and write conversation summaries without a Zapier subscription. The threshold: if you're syncing fewer than 200 contacts a month, the manual API calls cost more dev time than a no-code tool would. This MCP makes sense when you're already writing integration code, your CRM has a weird schema that breaks templated connectors, or you need sub-5-minute sync latency. For a team closing 40+ deals a quarter with custom contact attributes, the 10-tool surface area pays off in week two.

When paginated conversation export beats manual download

A fintech startup's compliance officer needs to pull 6 months of customer chat logs for a quarterly audit, filtered by date and tagged by topic. Insighto's conversation-list tool lets you script the export with date ranges and filters instead of clicking through a web UI 40 times. The catch: if your audit cadence is annual and you have fewer than 500 conversations, the one-time setup cost (API key, pagination logic, CSV formatting) outweighs the 20 minutes of manual clicking. This MCP wins when you're exporting logs monthly, need to cross-reference conversation IDs with contact records, or you're handing the export to a non-technical auditor who wants a repeatable script. For quarterly reviews at 1,000+ conversations, you'll recover the setup time in the first run.