Articles, guides and field reports on AI agents
Jev vs LLM-as-a-judge: we measured 4,500 judgements
Can Jev, TypeSafe's decision model, replace an LLM to evaluate your AI agents? We measured it on RM-Bench: accuracy, bias, calibration, cost.
Dust AI: what reviews miss about how it is actually used
Dust AI explained: features, use cases by team, pricing model and the one usage signal that published reviews almost always ignore.
Claude Cowork: capabilities, limits and how to manage it in the enterprise
Where Claude Cowork fits in your AI estate: what it can do, where it falls short and what to measure before you roll it out across a team.
Artificial intelligence in business: adoption, costs and real ROI
Measure the real ROI of artificial intelligence in business: use cases by function, a five-step deployment method and the key risks to plan for.
ChatGPT Enterprise: measure adoption before you decide
Measure real ChatGPT Enterprise adoption before you invest: what the plan includes, how the contract works, how secure it is, and how to decide.
AI in HR: measure real usage to know what to industrialise
Measure how AI is really used in HR across your company, so you can decide which HR AI use cases to industrialise, consolidate or cut.
Claude Code Harness: 98.4% Infrastructure, 1.6% AI
Configure your Claude Code harness: core components, 3 autonomy patterns, real costs and the path to enterprise governance for AI agents.
Open source AI agent: compare, choose and deploy to production
Compare 6 open source AI agent frameworks (LangGraph, CrewAI, AutoGen): enterprise criteria, licences and the move to production without lock-in.
Shadow AI: a need your company's AI offer does not cover
If your teams use their personal ChatGPT, your company is not giving them the AI they need. What shadow AI is, and how to make it recede.