AI Sales Plays

Today’s plays › All plays › Prospecting & research

Play · Prospecting & research
For sellersFor sales leadersGong call recordings, Relevance AI (agent hosting)

Before you build an outbound agent, mine your call recordings for why buyers took the meeting, and review the agent like an employee every week

Picked and checked by Olga Pearson, SalesUpLevel

Get tomorrow’s play in your inbox. Free ↓

The resultas reported

Reported in a RevGenius interview (Aug 22, 2026): in Q1, Angelo Quiambao's solo outbound agent generated 190 meetings and $15.4M in sourced pipeline, and a team co-pilot agent generated 132 meetings and $13.7M, for $29.2M combined. He built these agents at his previous company, before his current role at Intellistack. These totals are for the whole agent setup, not the trigger step alone.

Everyone buys the same signals: job changes, funding, hiring. The trigger nobody else can copy is sitting in your own call recordings.

Angelo Quiambao is the AI GTM engineer at Intellistack (formerly Formstack); he built the agents behind these numbers at his previous company, and says he no longer has access to them. He says he watched over 500 Gong calls and noticed that in the first 10 minutes buyers say why they took the meeting: a competitor that is not working out, an AI mandate, a pain that differs by persona. He treats those reasons as triggers and builds battle cards, value statements and intros from them. He scaled outbound steps his human team had already proven, not new ones, and kept a person approving anything customer-facing. Every week he asks each agent what it did well, what it did not, and how it tracked against its goal, and keeps a memory of the emails that led to meetings so the agent can repeat them. He now has a separate agent check the others daily and give each a health score.

By Angelo Quiambao (Intellistack), on the RevGenius Podcast, AI go-to-market engineer at Intellistack; built the agents behind these results at his previous company; former BDR at Qualified · YouTube: The 7 Layers of an Agentic GTM Motion | Angelo Quiambao, Intellistack | RevGenius Podcast (write-up: revenuecreator.com) ↗ · Aug 22, 2026

Run it before your 10am

  1. Pull 20 to 50 recorded first calls that came from outbound.
  2. Have AI list, for each call, the reason the buyer gave for taking the meeting in the first 10 minutes. Watch a sample yourself to check it read them right.
  3. Group the reasons by persona (executive vs. the person who feels the pain day to day) and count them.
  4. Turn the top reasons into the agent's triggers, openers and battle cards, using only outbound steps your team has already proven convert.
  5. Keep a person approving every customer-facing message.
  6. Every week, ask the agent what worked, what did not and how it did against its meeting goal; save the emails that booked meetings as examples it must study.
Copy-paste prompt · Our prompt (the interview does not share Angelo's)
Here are transcripts of first sales calls that came from outbound: [paste, or attach].
For each call, read only the first 10 minutes and answer:
1) Why did the buyer say they took this meeting? Quote them.
2) Their role, and whether this is an executive-level pain or a day-to-day user pain.
3) Was this really outbound, or inbound, partner or renewal? Skip anything that is not net-new outbound.
Then group the reasons, count how often each appears, and write one opening line per top reason that names that pain without guessing anything we cannot prove.

If you lead a team

  1. Give the agent a goal like a seller's, and review it weekly like a seller.
  2. Only automate outbound steps your people have already proven.

Watch out for

  • The meeting and pipeline totals are for two whole agents over one quarter, as stated by the host and confirmed by Angelo; the trigger step is not measured on its own, and the quarter's year is not stated.
  • Sourced pipeline is not closed revenue.
  • He stresses it worked because the human outbound motion was already converting; automating a broken process gives you a broken process at volume.
  • In the same interview he says he no longer has access to those agents because he is at a new company.
  • AI can misread why a buyer took the call; he watches the calls himself to check.
  • What I'd have my team try first: run the prompt on 20 outbound first calls and compare the top reasons with what your current sequences actually say.

Source: RevGenius Podcast with Angelo Quiambao, YouTube, uploaded Aug 22, 2026; Revenue Creator newsletter write-up by Jared Robin, Aug 22, 2026 ('The Weapon is already in your CRM'). Numbers and quotes checked against the video's captions. Results are as reported; we did not measure them.

Run these next

Live cohort · SalesUpLevel

The AI-Native Seller

An 8-week live cohort, Wednesdays, starting Nov 11, 2026. Turn plays like these into how you and your team sell, every week.

Get in the room
Coming soon

Hiring AI-native sellers?

For sales leaders building a team that already works this way. More soon.