sales-callminer

What this skill does

# CallMiner Platform Help

Help the user with CallMiner Eureka platform questions — from interaction capture and automated QA through agent coaching, real-time alerts, compliance monitoring, and API integration.

## Step 1 — Gather context

If `references/learnings.md` exists, read it first for accumulated platform knowledge.

Ask the user:

1. **What area of CallMiner do you need help with?**
   - A) Capture — recording, screen capture, redaction setup
   - B) Analyze — categories, scoring, sentiment, topic extraction
   - C) Coach — agent scorecards, coaching plans, performance tracking
   - D) RealTime — live alerts, next-best-action, real-time guidance
   - E) Compliance — PCI/HIPAA monitoring, PII redaction, risk flagging
   - F) API — Ingestion API (audio/text in), Data API (insights out)
   - G) Integrations — CRM, CCaaS, WFM, transcription engine
   - H) OmniAgent / Outreach / LiveTranslate — automation modules
   - I) Admin — licensing, user management, permissions
   - J) Something else

2. **What's your role?**
   - A) QA Manager / QA Analyst
   - B) Contact center manager / supervisor
   - C) Compliance / risk officer
   - D) CX / VoC analyst
   - E) IT / admin / developer
   - F) Other

3. **What are you trying to accomplish?** (describe your specific goal or issue)

**If the user's request already provides most of this context, skip directly to the relevant step.** Lead with your best-effort answer using reasonable assumptions (stated explicitly), then ask only the most critical 1-2 clarifying questions at the end.

## Step 2 — Route or answer directly

| Problem domain | Route to |
|---|---|
| CCaaS platform comparison/selection | `/sales-ccaas-selection {question}` |
| Agent coaching strategy (not CallMiner-specific) | `/sales-coaching {question}` |
| Customer feedback / NPS / CSAT strategy | `/sales-customer-feedback {question}` |
| Connecting CallMiner to other tools (architecture) | `/sales-integration {question}` |

Otherwise, answer directly from platform knowledge using the reference below.

## Step 3 — CallMiner platform reference

**Read `references/platform-guide.md`** for the full platform reference — modules, pricing, integrations, data model, API overview, workflows.

Answer the user's question using only the relevant section. Don't dump the full reference.

## Step 4 — Actionable guidance

You no longer need the platform guide — focus on the user's specific situation.

1. **Step-by-step instructions** for their goal in CallMiner
2. **Configuration recommendations** — specific settings, category rules, scoring criteria
3. **Common pitfalls** — what goes wrong and how to avoid it
4. **Verification** — how to confirm the change worked
5. **For API questions** — point to `references/callminer-api-reference.md`

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.*

- **Expect a 3-6 month ramp-up.** CallMiner is powerful but complex — building accurate categories, scoring models, and coaching workflows takes months. Don't expect instant ROI. Budget for dedicated analyst time during setup.
- **Auto-logout after 30 minutes is by design.** The session timeout is short and affects all open tabs simultaneously. Save work frequently. This is a known pain point with no user-configurable override.
- **Category tuning is iterative, not one-shot.** Initial category rules will have false positives/negatives. Plan for weekly refinement cycles during the first quarter. Test categories against known-good interactions before deploying to scoring.
- **Transcription engine choice affects everything downstream.** CallMiner supports multiple ASR engines (Deepgram, Google, Azure, Nuance) via OVTS. Accuracy varies by accent, industry jargon, and audio quality. Test multiple engines with your actual call recordings before committing.
- **Pricing is opaque — negotiate hard.** No public pricing. Average ~$102K/year but ranges widely. Seat-based vs. hours-analyzed licensing models have very different economics depending on your volume and agent count.

- **Self-improving**: If you discover something not covered here, append it to `references/learnings.md` with today's date.

## Related skills

- `/sales-coaching` — Sales coaching, QA, and agent training strategy (platform-agnostic)
- `/sales-ccaas-selection` — Compare CCaaS platforms (Genesys, NICE, Talkdesk, Five9, etc.)
- `/sales-observe-ai` — Observe.AI — contact center AI for QA and agent assist (CallMiner alternative)
- `/sales-cresta` — Cresta — enterprise contact center AI (CallMiner alternative)
- `/sales-balto` — Balto — real-time agent assist (CallMiner RealTime alternative)
- `/sales-convin` — Convin — conversation intelligence with auto QA
- `/sales-enthu` — Enthu.AI — contact center conversation intelligence
- `/sales-customer-feedback` — Customer feedback, NPS, CSAT, VoC strategy
- `/sales-integration` — Connect CallMiner to CRM, CCaaS, or other tools
- `/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 -a claude-code -y`

## Examples

### Example 1: Automated QA scoring setup
**User says**: "We're only scoring 2% of calls manually — how do I set up automated QA in CallMiner?"
**Skill does**:
1. Reads platform guide for Analyze and Coach modules
2. Explains category-based scoring — define quality criteria as categories, assign weights, auto-score 100% of interactions
3. Walks through creating a QA scorecard with weighted categories (compliance, empathy, resolution, script adherence)
4. Recommends starting with a calibration period comparing auto-scores to manual scores
**Result**: User has a plan to move from 2% manual sampling to 100% automated QA scoring

### Example 2: Compliance monitoring
**User says**: "We need to monitor all calls for PCI compliance — agents sometimes read back full card numbers"
**Skill does**:
1. Reads platform guide for compliance and Redact modules
2. Explains how to create categories detecting PCI-sensitive language patterns
3. Walks through Redact configuration for automatic PII masking in transcripts and audio
4. Recommends RealTime alerts for live intervention when PCI violations detected
**Result**: User has PCI compliance monitoring with auto-redaction and real-time agent alerts

## Troubleshooting

### Steep learning curve — don't know where to start
**Symptom**: Platform feels overwhelming, too many modules and configuration options
**Cause**: CallMiner is enterprise-grade — it's designed for dedicated analysts, not casual users
**Solution**: Start with ONE use case (e.g., automated QA on a single queue). Build 5-10 categories, validate against known interactions, then expand. CallMiner offers a sandbox environment for testing. Budget 3-6 months for the first production deployment. Request CallMiner's onboarding support — initial hand-holding is available.

### Categories not matching expected interactions
**Symptom**: Auto-scoring misses interactions that should match, or flags ones that shouldn't
**Cause**: Category rules are too narrow (missing synonyms/variations) or too broad (matching unrelated phrases)
**Solution**: Review the category's keyword/phrase rules. Use CallMiner's "Test" feature to run categories against a sample set. Add phrase variations and exclusions iteratively. Check transcription accuracy first — poor ASR = poor category matching regardless of rule quality.

### Pages timing out / session drops
**Symptom**: UI tabs time out and log you out across all open pages simultaneously
**Cause**: 30-minute session timeout is system-wide; one tab expiring kills all sessions
**Solution**: Save work frequently. Avoid having many tabs open simultaneously. If running long analysis queries, use the API (Data API) for batch exports inste