Claude for RevOps

AI in 2026: The conversation has changed

The conversation has shifted. Six months ago teams were asking whether to use AI. Now the question is how fast they can deploy it.

Jeff Ignacio, Founder of RevOps Impact, and Hassan Irshad, Head of RevOps at Unify, both pointed to the same inflection: speed of deployment is the new table stakes, build vs. buy is tipping toward build even at companies without engineering teams, and two months is genuinely an eternity in this space.

How RevOps is evolving with AI

RevOps has always been under-resourced relative to its impact. AI changes that dynamic.

Both panelists framed the shift the same way: AI gives RevOps a world-class analyst to delegate tactical work to, freeing up time to operate strategically. The function does not need to stay stuck in ticket queues and manual reporting. It can become the strategic layer it was always supposed to be.

What you need in place before you deploy

Not every model is created equal and not every use case needs an expensive one. Jeff walked through the core prerequisites:

• Match the model to the task
• Architect your AI differently than you use the chat interface
• Invest in prompt engineering and context window management
• Build clean data foundations first -- garbage in, garbage out

Hassan reinforced the foundation piece. The quality of your schema, fields, and processes determines what AI can do with them. Layer it on solid foundations and use cases compound. Layer it on chaos and you get chaos back.

Use cases in action

Hassan demoed a closed-lost analysis agent he built in Cowork. It pulls CRM data, call transcripts from Attention, and Slack conversations, then cross-references all of it to surface loss reasons the existing dropdowns were not capturing. What used to take hours now runs on a weekly schedule and posts directly to Slack.

He also showed a forecast slippage agent that auto-builds a postmortem on any deal that slips from commit -- pulling push counts, call history, SPICED notes, and calendar data -- and delivers it to Slack before the leadership forecast call.

Jeff shared two examples built in Claude Code: a territory scoring model that estimated account value across a full company population, and a sales rep scorecard that analyzed thousands of calls against a MEDDIC rubric to answer second- and third-order questions about rep performance over time.

Highest ROI use cases

The panelists pointed to three areas where AI delivers the clearest return:

• Signal synthesis: Turning noisy intent data into weighted, actionable triggers
• Pipeline reality: Using call intelligence to assess whether deals match their stage
• Recurring leadership deliverables: Collapsing hours of prep into minutes

The problems are not new. The solutions are.

Building a team of agents as a solo IC

The session closed on a question from the audience: how do you scale your impact without headcount?

Hassan's answer: start with yourself. Find what consumes the most time, build agents to handle it, and use the time you get back to operate at a higher level. One IC doing the work of four is how you prove the ROI that gets noticed.

Jeff added: either deploy AI where you are strong to free up time elsewhere, or use it to cover your gaps. Get read-only access across your GTM stack and you can do analysis you never had time for. Add write access (in a sandbox first) and the agents can take action too.