Building the Measurement Layer for an Agent-Run Web, A Conversation with Cameron Witkowski

2026-06-22 · Source: The AI Journal · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, E-commerce & Digital Commerce · Depth: Intermediate, long

Summary

OpenLens, co-founded by Cameron Witkowski, serves as an AI visibility platform for over 35 marketing agencies, measuring how brands appear across major AI models. Witkowski, previously from AWS AI Labs, emphasizes the critical gap between agent demos and production, identifying observability as the primary constraint. He notes that current web infrastructure, optimized for human users, often fails under agent traffic due to misconfigurations like blocked crawlers. OpenLens aims to build infrastructure for an agent-mediated web, addressing challenges such as attributing agent-driven conversions and preparing sites for non-human browsing patterns. Witkowski views this shift as a new economic engine, comparable to the rise of e-commerce, and highlights the need for new primitives for agent identity, payments, and trust. He is closely monitoring the adoption of WebMCP, a W3C draft co-authored by Microsoft and Google, which he believes will make agents first-class citizens on the web.

Key takeaway

For AI Engineers and Product Managers building web-facing applications, you must prioritize observability and agent-centric design from the outset. The web is evolving beyond human-only traffic, demanding infrastructure that supports agent interaction and measurement. Your systems need robust instrumentation to manage stochastic agent behaviors and adapt to emerging protocols like WebMCP. Failing to prepare your digital presence for agent traffic risks losing significant future economic opportunities.

Key insights

The web is rapidly shifting from human-centric to agent-mediated, requiring new infrastructure and observability for brands and commerce.

Principles

• Observability is crucial for controlling non-deterministic agent systems.
• Production systems should be hybrids of autonomous and deterministic logic.
• Don't hardcode model versions; assume API changes.

In practice

• Review Cloudflare configurations for agent crawler access.
• Build APIs and MCP access for agent interaction with dashboards.
• Implement CI/CD and monitoring for model version changes.

Topics

• AI Agents
• Agent-mediated Commerce
• Web Observability
• WebMCP Protocol
• Marketing Technology
• LLM Operations

Best for: CTO, VP of Engineering/Data, Executive, AI Engineer, MLOps Engineer, Director of AI/ML

Related on AIssential

• Building Agents at Scale: Lessons from the Front Lines With Gary Stafford — Agentic systems leverage LLMs, tools, and reasoning loops for non-deterministic problem-solving and automation.
• Towards an Agent-First Web: Redesigning the Web for AI Agents — The web needs a fundamental redesign to accommodate AI agents as first-class citizens, moving beyond human-centric assumptions.
• WebMCP - Why is awesome & How to use it — Web MCP enables deterministic AI agent interaction with websites through contextually loaded, declared actions.
• Long-Running AI Agent Browser Automation Tasks Is Here — AI agents demonstrate high persistence and adaptability in achieving complex, open-ended goals through browser automation.
• Something big is happening — Agentic AI is rapidly evolving, shifting software development towards intent-driven outcomes and human-agent collaboration.
• The Truth About Agents in Production — Effective AI agent development prioritizes type safety, robust evaluation, and a problem-first approach over complex multi-agent systems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The AI Journal.