Avoma

@avoma list all customer calls from the past week and show me which ones mentioned pricing or contract terms

Avoma returned 14 external calls from the past 7 days. Of these, 9 contained mentions of pricing or contract terms based on Avoma's topic extraction. The highest-engagement call was a 42-minute discovery session on March 18 with Acme Corp, where pricing came up in three separate segments totaling 8 minutes. Two calls from March 20 flagged contract renewal discussions. This example illustrates how Avoma's list and analysis endpoints surface conversational themes across a date range.

This example combines list filtering with AI topic analysis. Requires read:meetings scope. Note that Avoma processes recordings asynchronously — calls uploaded in the last few hours may not yet have analysis available. Topic extraction quality depends on audio clarity and Avoma's training data for your domain.

@avoma create a call record for the Zoom meeting I just hosted with the recording URL from our storage, then confirm it's queued for transcription

Avoma accepted the external call submission. The recording URL you provided is now queued for download and processing. Avoma assigned external ID ext-2024-03-21-acme-demo and will begin transcription once the file is retrieved. Processing typically completes within 15–30 minutes depending on recording length. You can poll the call's readiness state using the returned ID. This example shows how Avoma ingests externally hosted recordings rather than requiring native integration.

This write operation creates persistent state in Avoma. Requires write:calls scope. The recording URL must be publicly reachable by Avoma's servers — pre-signed S3 URLs or equivalent work well. Be aware that submitting a call triggers billable processing; confirm your team's Avoma plan limits before automating bulk uploads.

@avoma pull the AI analysis for meeting UUID abc-123, then draft a follow-up email highlighting the three most important action items and any unresolved objections

Avoma's analysis identified three action items: (1) send pricing proposal by March 25, (2) schedule technical deep-dive with engineering lead, (3) provide case study for healthcare vertical. One unresolved objection surfaced: concern about data residency for EU customers, discussed at 18:32 in the recording. Draft email: 'Hi [Name], thank you for the productive conversation. Based on our discussion, I'll send over pricing by Friday and coordinate a technical session with our engineering team. I've also noted your question about EU data residency — let me loop in our compliance lead to address that directly.' This example shows how Avoma's structured insights feed directly into post-call workflows.

This synthesis example pairs Avoma's meeting analysis endpoint with the AI's summarization. Requires read:meetings scope. Analysis is only available after Avoma finishes processing the recording. Sentiment and topic segmentation accuracy improves with longer, structured conversations — brief or chaotic calls may yield sparse insights.

When Avoma beats manual call notes for weekly pipeline review

A sales team of eight reps runs a Friday pipeline review where the manager spot-checks three calls from each rep's week. Without Avoma, the manager scrubs through recordings or reads inconsistent notes. With the MCP, the manager queries Get Meeting Analysis for sentiment and topic breakdowns across all 24 calls in under a minute, then pulls Get Recording URLs for the two outliers that need a full listen. The trade-off: if your team already lives in Gong or Chorus and those platforms feed your CRM, adding Avoma creates a second source of truth. But if you're on a lighter stack or need async processing of external recordings (webinars, support calls), the Create Call and List Calls tools let non-sales teams push recordings into the same analysis pipeline. This works when your call volume stays under 200 per week and you value a single AI lens across customer conversations.

Using engagement metrics to catch onboarding drift

A four-person CS team runs 60 onboarding calls a month using a standard 45-minute playbook. Three months in, renewal rates dip and no one knows why. The team uses Get Engagement Summary filtered by meeting type and date range to surface aggregate talk ratios and question counts, then spots that two CSMs are monologuing for 35 of the 45 minutes while the other two hold closer to 20. They pull Get Meeting Analysis on a sample from each CSM to confirm the pattern, then retrain. The constraint: Avoma's engagement metrics are retrospective, not real-time, so this is a monthly or quarterly audit play, not an in-call coaching tool. If you need live nudges during calls, you want a different MCP. But for small CS or sales-eng teams that want to measure playbook adherence without a full QA hire, the 16 tools here cover meeting ingest, analysis, and aggregate reporting in one integration.

When Avoma's topic segmentation speeds qual research

A product trio (PM, designer, researcher) runs 12 user interviews over two weeks to validate a feature concept. Traditionally, the researcher transcribes and tags themes in a spreadsheet over three days. With the MCP, they upload recordings via Create Call, wait for async processing, then use Get Meeting Analysis to pull Avoma's topic segments and sentiment tags across all 12 sessions in one batch query. The PM filters for mentions of the target feature and exports a CSV of timestamps and sentiment scores, cutting synthesis time to under four hours. The limitation: Avoma's topic extraction is tuned for sales and CS language, so highly technical or domain-specific interviews may need manual re-tagging. If your research vocabulary is standard (pricing, onboarding, pain points), this works. If you're interviewing ML engineers about tensor optimization, expect to supplement with your own tagging. Best fit for teams running 10-30 interviews per quarter who want faster pattern-spotting without hiring a research ops person.