Skip to main content

Usage Examples

These examples show real conversations between users and AI assistants connected to AiSky via MCP. Each example shows the natural language prompt, the tools the assistant calls, and the result.
These examples work with any MCP-compatible client, Claude Desktop, Claude Code, Cursor, VS Code, or ChatGPT. The AI assistant automatically selects the right tools based on your request.

Example 1: Contact Management Workflow

Import and organize contacts

You: “Add a new contact Sarah Chen, email sarah@acmerealty.com, phone +1 415 555 0123, company Acme Realty. Tag her as enterprise and buyer.” Assistant calls: smartalex_create_contact
Result:
You: “Now create a R75,000 deal for Sarah called ‘Acme Realty Enterprise’ in the qualified stage.” Assistant calls: smartalex_create_deal
Assistant: “Done. Created the ‘Acme Realty Enterprise’ deal for R75,000 in the qualified stage, linked to Sarah Chen.”

Example 2: Call Analysis and Follow-Up

Review recent calls and take action

You: “Show me my last 5 calls.” Assistant calls: smartalex_list_calls with per_page: 5 Assistant: “Here are your 5 most recent calls: Call #2 with Mike Rodriguez had a low sentiment score. Want me to pull the transcript?” You: “Yes, show me Mike’s call transcript.” Assistant calls: smartalex_get_call with the call UUID Assistant: “Here’s the transcript from Mike’s call. He expressed frustration about scheduling delays. Key objections: timing conflict with the Q2 deadline, and pricing concerns for the premium tier. I’d recommend a follow-up call with a discount offer or flexible scheduling options.” You: “Update Mike’s contact, add a tag ‘needs-follow-up’.” Assistant calls: smartalex_update_contact

Example 3: Pipeline Review and Deal Management

Analyze your sales pipeline

You: “Give me a pipeline overview. How many deals in each stage?” Assistant calls: smartalex_list_deals (all stages) Assistant: “Here’s your pipeline breakdown: Total pipeline: 19 deals worth R556,500. Your offer stage has the highest average deal value (R75,000). The showing-to-offer conversion looks strong, want me to break down which deals are closest to closing?” You: “Move the Acme Realty deal to the offer stage.” Assistant calls: smartalex_update_deal
Assistant: “Done. ‘Acme Realty Enterprise’ moved from qualified to offer stage.”

Example 4: Full Campaign Workflow

Launch an outbound campaign end-to-end

You: “I want to set up an outbound campaign.” Assistant calls: smartalex_list_agents and smartalex_list_phone_numbers Assistant: “Here’s what’s available: Agents: Phone Numbers: Which agent would you like to use, and what should the campaign objective be?” You: “Use the Sales Assistant. The objective is to schedule product demos with enterprise leads. Use the US number and set it to call weekdays 9am to 5pm Eastern.” Assistant calls: smartalex_create_campaign
Assistant: “Created ‘Q2 Enterprise Outreach’. Now let’s add contacts. Which contacts should I include?” You: “Add Sarah Johnson and Mike Chen.” Assistant calls: smartalex_add_contacts_to_campaign
Assistant: “Added 2 contacts. The campaign is ready to launch, want me to start it?” You: “Yes, start it.” Assistant calls: smartalex_start_campaign
Assistant: “Campaign ‘Q2 Enterprise Outreach’ is now active. The Sales Assistant will begin calling 2 contacts during your configured business hours (Mon, Fri 9am, 5pm ET).”

Example 5: Webhook Integration

Set up real-time event notifications

You: “I want to send call completion events to my CRM at https://crm.example.com/webhooks/smartalex Assistant calls: smartalex_create_webhook
Result:
The signing_secret is shown only once. Save it immediately to verify webhook payloads on your server using HMAC-SHA256. The verification pattern is the same one documented in HTTP Tools signature verification, code samples for Node, Python, PHP, and Go are there.
You: “Show me all my active webhooks.” Assistant calls: smartalex_list_webhooks Assistant: “You have 2 active webhooks: Both are healthy with zero failures.”

Example 6: Platform Health Check

Daily operations check

You: “How’s my AiSky platform doing?” Assistant calls: smartalex_get_platform_status Assistant: “AiSky is operational. Here’s your account overview: Everything looks healthy. Your call volume is up 12% from last week.”

Multi-Step Workflows

The AI assistant can chain multiple tools together in a single conversation to complete complex workflows.

Lead qualification workflow

You: “Run a pipeline review. Show me my qualified leads, check if any have recent calls with negative sentiment, and flag deals that might need attention.” The assistant will automatically:
  1. Call smartalex_list_deals with stage: "qualified"
  2. For each deal with a contact, call smartalex_list_calls to check recent activity
  3. Cross-reference sentiment scores
  4. Provide a summary with recommendations

Campaign performance analysis

You: “How did my Q2 Enterprise Outreach campaign perform? Show me the calls, sentiment breakdown, and any deals that came from it.” The assistant will:
  1. Call smartalex_list_campaigns to find the campaign
  2. Call smartalex_list_calls to pull calls from that campaign
  3. Call smartalex_list_deals to find related deals
  4. Provide conversion analysis and ROI estimates

Using Prompts

AiSky includes 5 built-in prompts that guide the AI through complex analysis:

Campaign Strategy

You: “Use the campaign-strategy prompt to help me design a follow-up campaign for my real estate leads.” The assistant will:
  • Pull your current contacts, agents, and campaigns
  • Analyze which contacts haven’t been reached
  • Recommend agent assignment, timing, and messaging
  • Draft a campaign plan

Pipeline Review

You: “Run a pipeline review.” The assistant uses the pipeline-review prompt to:
  • Analyze deals at every stage
  • Identify bottlenecks (stages with stale deals)
  • Flag at-risk deals (high value, no recent activity)
  • Provide a forecast based on conversion rates

Lead Qualification

You: “Score my leads using the lead-qualification prompt.” The assistant will:
  • Pull contacts tagged as leads
  • Check call history and transcript sentiment
  • Score each lead on engagement, fit, and timing
  • Rank leads by likelihood to convert

Tips

Instead of “update the deal”, say “move the Acme Realty deal to offer stage”, the assistant will search by name.
You can say “Create a contact, then create a deal for them” in one message. The assistant handles the sequencing.
Instead of “list my calls”, try “which calls had the lowest sentiment this week?”, the assistant will filter and analyze for you.
When you need multi-step analysis, reference a prompt by name: “Use the pipeline-review prompt” or “Run a lead qualification analysis.”

Next Steps