Step 1 — Gather context
If references/learnings.md exists, read it first for accumulated platform knowledge.
- What do you need?
- A) Set up or refine QA scorecards
- B) Build coaching workflows from QA data
- C) Configure conversation analytics / AskAI
- D) Integrate with helpdesk, phone system, or CRM
- E) Export data to warehouse or via API
- F) Monitor AI chatbot quality
- G) Ingest CSAT scores and correlate with QA
- H) Other
-
Current setup?
- A) New to MaestroQA (not yet connected)
- B) Connected to helpdesk, building scorecards
- C) Running QA, want to improve coaching
- D) Mature — need analytics, API, or warehouse export
-
Helpdesk / phone system?
- A) Zendesk / Freshdesk / Intercom / Front / Gorgias
- B) Salesforce Service Cloud / ServiceNow / Dynamics 365
- C) Five9 / Talkdesk / Amazon Connect / NICE / Genesys
- D) Aircall / RingCentral / Dialpad / Twilio / Vonage
- E) Other / multiple
Skip-ahead rule: if the user's prompt already contains enough context, skip to Step 2.
Step 2 — Route or answer directly
| Problem domain | Route to |
|---|---|
| Comparing QA tools (MaestroQA vs Observe.AI vs Enthu.AI) | /sales-coaching for cross-platform comparison |
| Choosing a CCaaS platform | /sales-ccaas-selection |
| Real-time agent coaching during calls | /sales-balto or /sales-cresta |
| Help desk platform selection | /sales-helpdesk-selection |
| Salesforce-specific questions | /sales-salesforce |
| Sales call review and coaching | /sales-call-review |
When routing, provide the exact command: "This is a {problem domain} question — run: /sales-{skill} {user's original question}"
Step 3 — MaestroQA platform reference
Read references/platform-guide.md for the full platform reference — modules, pricing, integrations, data model, workflows.
Answer the user's question using only the relevant section. Don't dump the full reference.
Step 4 — Actionable guidance
Focus on the user's specific situation.
- Scorecard design: Start with 5-8 binary criteria per scorecard. Binary (yes/no) scores faster and more consistently than scaled (1-5). Add weighted sections for compliance vs quality vs sales effectiveness.
- Coaching connection: Link low QA scores to coaching sessions automatically — don't let scores sit in reports. Use reverse-ETL to push coaching tasks to Slack or your CRM.
- Analytics adoption: Start with AskAI for ad-hoc queries before building custom dashboards. AskAI answers natural language questions about conversation data without needing SQL.
- API automation: Use the Rippit API for bulk data export, CSAT ingestion, and agent provisioning. Rate limit is 10 req/s — batch operations where possible.
- Chatbot QA: Connect Ada, Sierra, Agentforce, or Forethought to grade bot conversations with the same scorecards used for human agents.
If you discover a gotcha, workaround, or tip not covered in references/learnings.md, append it there.
Gotchas
Best-effort from research — review these, especially items about plan-gated features and integration gotchas that may be outdated.
- AI features cost extra. MaestroQA's AI-powered analytics and conversation intelligence are add-ons — budget separately from the base QA subscription.
- Ticket sync delay. Zendesk tickets typically take 1-3 hours to sync. By default only tickets updated within the last 30 days are kept in the app database. Plan QA workflows around this lag.
- API tokens expire every 90 days by default. Set up a rotation process or disable expiry for long-lived integrations.
- Export format limitations. Some users report inability to export grading progress as CSV/PDF directly from the UI. Use the Rippit API
POST /request-raw-exportendpoint as a workaround. - Dashboard customization is limited. For advanced analytics, export to Snowflake/BigQuery and use external BI tools.
- Learning curve. MaestroQA requires upfront investment to configure scorecards, calibration sessions, and coaching workflows. Budget 2-4 weeks for initial setup with a dedicated QA admin.
- Rate limits. API is 10 req/s and 100 req/min. Batch operations (comments, metrics) accept up to 500 items per request — use these instead of individual calls.
Related skills
/sales-coaching— Sales coaching and training strategy across all QA and enablement platforms/sales-observe-ai— Observe.AI platform help (enterprise contact center QA with Auto QA, Agent Copilot)/sales-enthu— Enthu.AI platform help (affordable contact center QA, fast setup)/sales-balto— Balto platform help (real-time AI guidance during calls)/sales-cresta— Cresta platform help (enterprise contact center AI)/sales-convin— Convin platform help (contact center QA + coaching + LMS)/sales-ccaas-selection— Choosing a CCaaS platform/sales-helpdesk-selection— Choosing a help desk platform/sales-do— Not sure which skill to use? The router matches any sales objective to the right skill. Install:npx skills add sales-skills/sales --skill sales-do
Examples
Example 1: Set up QA scorecards for a support team
User says: "We just connected Zendesk — how do I build QA scorecards for our support team?" Skill does: Walks through scorecard design (binary criteria, weighted sections, compliance vs quality), calibration setup for evaluator consistency, and assignment rules for random sampling vs targeted review.
Example 2: Export QA data to Snowflake
User says: "I need to get our grading data into Snowflake for custom reporting"
Skill does: Covers the native Snowflake integration setup and the Rippit API export endpoints (POST /request-raw-export, GET /export-data/:id) as alternatives, with rate limit considerations.
Example 3: Correlate CSAT with QA scores
User says: "We want to see if our QA scores actually predict customer satisfaction" Skill does: Explains the CSAT Bulk Ingestion API, how to connect Qualtrics/Delighted/Simplesat for native CSAT ingestion, and how to use AskAI to query correlations between QA scores and CSAT data.
Troubleshooting
QA scores don't correlate with customer satisfaction
Cause: Scorecards measure process compliance (did the agent follow the script) but not outcome quality (was the customer actually helped) Solution: Add outcome-oriented criteria — resolution quality, customer effort reduction, empathy signals. Use CSAT ingestion to validate that QA criteria predict customer satisfaction. If they don't, revise criteria.
Evaluators score the same ticket differently
Cause: Criteria are ambiguous or subjective Solution: Use binary (yes/no) criteria instead of scaled (1-5). Run calibration sessions monthly — have 3+ evaluators score the same 5 tickets independently, then discuss disagreements. MaestroQA has built-in calibration tools for this.
Coaching sessions feel disconnected from QA data
Cause: QA scores sit in MaestroQA dashboards but coaching happens in a separate tool or ad-hoc Solution: Use MaestroQA's coaching workflows — low scores automatically trigger coaching sessions with specific call segments linked. Use reverse-ETL to push coaching tasks to Slack or your CRM so supervisors see them in their workflow.