How G2 Names and Classifies CX AI Agents
This note lays out the four ways companies get a CX agent today, what we call them, and what to weigh before you choose one.
Build Notes.
What we’re building, testing, and learning across G2’s AI work.
RSS feed6 Articles
Ruehl JohnsonLeaderboards
Introducing the G2 Investor Leaderboard
G2 now ranks 50 venture capital and private equity firms by how their portfolio companies perform on G2 and in AI answers. See where your firm stands.
Why we built Blueprints
Why we built Blueprints, how teams share AI workflows, what safety checks mean, and what we’re working on next.
How G2 compares foundation models with AI CX agents on the leaderboard
G2 compares 14 foundation models with AI CX agents on shared support tasks, explains the default model baselines, and reports outcomes and policy compliance.
How G2 evaluation scoring works for AI CX agents
How G2 scores CX agents on shared tasks, separates outcomes from supporting dimensions, and publishes the evidence behind each result.
G2 is Building the New Trust Layer for AI Agents
Why we’re building agent evaluations, what the first results revealed, and where the public beta goes next.