AI in sales masterclass & why RevOps will disappear with Kyle Vamvouris
10 snips
Apr 7, 2026 Kyle Vamvouris, founder of SalesThread and AI practitioner focused on conversational intelligence for sales. He discusses making AI simple for reps and building agents to extract structured call data. He explains scoring rubrics, objection detection, and how AI could replace routine deal support and reshape RevOps and management spans.
AI Snips
Chapters
Transcript
Episode notes
Widespread AI Adoption Often Lacks ROI
- Many companies deploy AI but fail to see ROI, with over half of CEOs reporting no AI ROI and 70–85% of sales AI deployments missing expectations.
- Kyle cites Forbes data and industry stats showing 81–90% of sales teams use AI yet most projects underperform, framing the ROI problem.
Built SalesThread From Years In The Trenches
- Kyle built SalesThread after repeatedly seeing AI sales tools made by people who hadn't sold and lacked a sales point of view.
- He leveraged his experience building 87 sales teams and coaching thousands to design AI that automates tasks between calls and provides actionable feedback.
Context Engineering Fixes Long Transcript Limits
- Long call transcripts (30–60k tokens) break LLM context windows, so querying raw transcripts yields poor answers without engineering.
- SalesThread extracts structured signals with agents (context engineering) to feed LLMs only the relevant summarized data, then fetches original text for gaps.
