How to Scale Customer Success with AI Agents ft. LeeRon Yahalomi (Aligned)

In this episode, LeeRon Yahalomi joined Aligned as VP of CS with two people and 100 accounts. Instead of hiring her way out of the problem, she built an army of AI agents in her first two weeks. What she built, how she built it, and why she thinks CS teams are still leaving most of their revenue on the table.

---

Listen on YouTube, Spotify and Apple Podcasts

---

Three Things We’re Taking Away:

1. Build agents like you build a team. LeeRon retrains her agents weekly based on what she liked and didn’t like from the prior week. She feeds them her DISC profile, her communication style, and sample examples of how she wants to sound. Agents aren’t set-and-forget. They’re direct reports that need calibration.

2. The morning prep agent is the highest-leverage place to start. Before her first meeting, LeeRon fires up ChatGPT, and her agents surface what she covered with each customer last time, open action items, and suggested talking points. Thirty minutes of prep, automated. The result: she walks into back-to-back calls fully present instead of side-swiveling mid-conversation.

3. CS is a revenue center, not a cost center. 58% of sales growth right now is driven by expansion, per an Emergence report. If CS owns expansion and you’re not investing in CS infrastructure, you’re not managing a cost center. You’re leaving the majority of your growth on the table.

🎯 The [Un]Churned Take: What a Two-Person CS Team Looks Like When It Actually Scales

LeeRon Yahalomi, VP of Customer Success at Aligned, joined the company eight weeks before this episode with one other CSM and 100 accounts to manage. The growth plan got blown past almost immediately. This conversation is a rare look at what it actually takes to build a CS function that scales not by hiring, but by building.

The Four Agents and What They Actually Do

LeeRon has four core agents, all currently running in ChatGPT, with plans to connect them via n8n once they’re proven. The first handles AE-to-CS handoffs: it ingests call transcripts, CRM notes, and public company data (news, 10-K reports, competitor positioning) and produces a briefing that lets her walk into a kickoff call with knowledge rather than a deck. “We show up to that first kickoff ready to success you into a better position than you were.”

The second is her morning prep agent. Connected to her calendar, fed with call recordings after each customer conversation, it generates a daily briefing: here’s who you’re talking to today, here’s what you covered last time, here’s what was left open, here’s where to pick up. “I walk in, even if it’s back to back, ready to jump straight in.”

The third is her communication agent, and it might be the most personal. She fed it her DISC personality profile, wrote out how she naturally communicates, gave it examples of writing styles she admires, and asked it to help her find the middle. “Sometimes I feel granola,” she said. Crunchy. Abrupt. On those days, the communication agent softens the draft before it goes out.

The fourth is the Orchestrator: the agent she’s still building toward. It monitors the platform for behavioral signals beyond login data, listens to calls for sentiment, flags missing or misaligned CRM data, and tries to surface the things a CSM missed while they were focused on the conversation. “It allows me to be in the CRM, in the call recording, in the conversation with you, in the platform, all at the same time, while I’m just me.”

The Human Review Line She Won’t Cross

LeeRon is clear about one thing: the agents don’t send without her. She reads everything before it goes out. Not because she doesn’t trust the output, but because she wants to stay in the loop on what’s being said on her behalf. “Maybe I’m a soft brownie today.” The judgment call about tone stays human. The research, the prep, the drafting, the flagging, all of that is the agents’ job.

This is also why she builds in ChatGPT before connecting to live systems. She tests the agent, finds the kinks, corrects the behavior, and only then starts thinking about platform integrations. Crawl before you connect.

CS Is the Next Sale

LeeRon’s closing argument is the kind of position that’s easy to agree with and hard to actually act on. CS isn’t post-sales. The sales funnel doesn’t end at close. It restarts. Every renewal, every expansion, every QBR is a re-close. “At every given moment, at every single interaction, CS has an obligation to continue the sale.” And with 58% of growth now coming from expansion, the math on under-investing in CS doesn’t hold up.

Wrapping Up

Two people, 100 accounts, and an army of agents she retrains every week. LeeRon isn’t waiting for the enterprise playbook. She’s writing a new one.

See you next week 🧠

---

Where to find LeeRon Yahalomi:

• LinkedIn: https://www.linkedin.com/in/leeron-yahalomi-1b066819/

---

Referenced in This Episode

• Aligned — The collaboration platform LeeRon works at, connecting sellers and buyers in shared deal rooms and mutual action plans. Used across sales, CS, and onboarding. aligned.io

• ChatGPT — LeeRon’s current platform for building and running all four of her agents. She builds and tests here before connecting to any live systems. chat.openai.com

• n8n — The workflow automation platform LeeRon plans to use to connect her agents to internal systems once they’re proven. n8n.io