AI agents

A useful business agent answers from your own content, knows its limits and never produces data that commits the company. These articles cover design, deployment, cost and guardrails.

AI agents · 9 min read

AI agents in customer service: what they handle, what they hand over

Scope, knowledge base, handover with context, committing data, transparency, measurement: how to design an AI agent for customer service that holds up.

Read the article

AI agents · 10 min read

Evaluating an AI agent: building and maintaining its test set

Real cases, answers validated by the business, hard and out of scope cases, criteria, replays and production monitoring: building an AI agent's test set.

Read the article

AI agents · 10 min read

Choosing a language model for your business: the criteria that matter

Quality on your own cases, real cost, latency, context, API or open weights, contract and continuity: the criteria for choosing a business language model.

Read the article

AI agents · 11 min read

Dots and GPT-6.1 Sol: what OpenAI's DevDay changes for businesses

OpenAI launches dots, always-on agents, and GPT-6.1 Sol. We read the official pages: availability in Europe, controls, pricing and the limits to know.

Read the article

AI agents · 10 min read

Claude Sonnet 5.5: which model for which work, and should you migrate

Claude Sonnet 5.5 was released on September 28: faster, fewer tokens, same price. What the numbers say, and our step-by-step method for migrating an agent.

Read the article

AI agents · 10 min read

RAG in business: getting an agent to answer from your documents

Getting an AI agent to answer from your documents: preparing the corpus, handling permissions, splitting, citing sources, evaluating, and what RAG will not fix.

Read the article

AI agents · 6 min read

This week in AI: 21 to 26 September 2026

Claude Opus 5.5, Gemini synthetic voices, Dataiku's agent inventory, a planned common standards body: this week's AI news and what it changes.

Read the article

AI agents · 7 min read

Kabaun Studio joins Anthropic's Claude Partner Network

Anthropic has approved Kabaun Studio into the Claude Partner Network: what the programme is, what it changes for our clients, and how to check a partner.

Read the article

AI agents · 12 min read

Claude Opus 5.5: what it changes for your agents in production

Opus 5.5 shipped on 22 September 2026: pricing, announced gaps, and above all what a cost cut does or does not change for an AI agent already in production.

Read the article

AI agents · 9 min read

AI at trade shows, in service of the visitor

Conversational agent, 3D map, live operations, human handoff: how AI answers visitors at trade shows, fairs, conferences and festivals.

Read the article

AI agents · 8 min read

What an AI agent costs in production: the line items

Tokens consumed, hosting, integration with existing tools, human oversight: the real cost items of a production AI agent, and how to reduce them.

Read the article

AI agents · 8 min read

Why an AI agent must never quote a price

A chatbot that invents a price binds the company. Our answer: closed list, server-side calculation, review before display.

Read the article

AI agents · 7 min read

Deploying an AI agent in your company, step by step

What happens between the decision and going live: scoping, collection, configuration, real-world testing and monitoring. No folklore.

Read the article