Summary
Dr. Sajjan Kanukolanu, VP of Global Operations and Strategy at Position2, presents an AI-native GTM (go-to-market) architecture built around a core premise: by the time a B2B buyer reaches your website, the decision is largely made. Citing Forrester 2026 research, he notes that 94% of buyers use generative AI as a primary research tool, 67% prefer a rep-free experience, and 80% of deals go to vendors already on the buyer’s pre-contact shortlist.
The talk introduces a three-layer architecture to close this gap. The first layer handles real-time visitor identification and intent signal collection — combining multiple identity resolution tools in parallel since no single source captures every visitor. The second layer drives personalized engagement: addressing returning visitors by name, surfacing relevant content based on behavioral signals, and triggering the right sales alerts without overwhelming reps. The third layer continuously updates a per-account context graph that maps individuals, buying committees, and deal-stage signals into a connected record the sales team can act on.
Position2 has deployed this architecture across 75+ client-specific AI agents powered by 18+ vertical knowledge bases, with over 800 agent runs per month as of June 2026. Kanukolanu’s central argument is that bolting AI onto existing GTM stacks addresses at most two of three problems (AI capability and integration) while leaving the underlying architectural model broken — and that all three must be solved simultaneously for meaningful scale. Relevant for revenue, marketing, and enterprise AI teams evaluating AI-native CRM and sales intelligence approaches.
📺 Source: AI Engineer · Published July 20, 2026
🏷️ Format: Workflow Case Study
