Skip to content
WorkPortfolioSolutionsBlogAboutContact
FrançaisFree diagnostic

Notes on AI that actually helps.

What we learn building AI tools for real businesses — written in plain language, with real examples. No hype, no names.

Business & compliance

13 September 2026·5 min read

The EU Cyber Resilience Act’s vulnerability-reporting duty is now live — and it may already apply to your software

On September 11, 2026, Article 14 of the EU Cyber Resilience Act took effect: any company that "manufactures" a product with digital elements — hardware or software sold under its own brand — must now report an actively exploited vulnerability within 24 hours and a severe incident within 72 hours, to ENISA and France’s national CSIRT, under fines of up to €15 million or 2.5% of global turnover. A 2024 academic survey found only 12.3% of SMEs were even aware the regulation existed. Here is what actually changed, who it concerns, and where to start.

Read →

Strategy, costs & ROI

11 September 2026·4 min read

OpenAI opens up its agent infrastructure: what the Agents API changes for build-vs-buy in your SMB

On September 10, 2026, OpenAI launched the Agents API in public beta, exposing the same "harness" that runs Codex — session management, orchestration, context compaction, recovery — behind a single API call. It makes agent infrastructure cheaper to assemble. It does not make an agent that reliably touches your invoicing, your CRM or your inbox easier to build without engineering supervision.

Read →

Business & compliance

10 September 2026·4 min read

CNIL fines Extia €300,000 over botched erasure requests: what it changes for any SMB that recruits

On September 9, 2026, the CNIL published a €300,000 fine against IT staffing firm Extia for mishandling 265 erasure requests from candidates and former employees in 2024 — more than three-quarters were not processed correctly. The ruling draws a sharp line most small recruiting teams miss: deleting the data is not enough if you never tell the person you did it.

Read →

Strategy, costs & ROI

8 September 2026·4 min read

OpenAI calls for slowing down AI after a Claude/ChatGPT/Grok outage — what it means for your AI agents

On September 6, 2026, OpenAI’s chief scientist published an essay calling for voluntary slowdowns in AI development, saying no lab can yet reliably monitor its most advanced models. It came three days after a multi-hour outage hit Claude, ChatGPT and Grok at the same time. Two separate signals, the same week, pointing at the same blind spot for any SME running AI agents without supervision or a fallback plan.

Read →

Studio notes

7 September 2026·4 min read

Studio notes — making AI agents reliable in production

This week the agent fleet shipped no big new feature, but three architectural choices spotted across ongoing engagements are worth sharing: caching AI email triage instead of recomputing it on every click, keeping a Qualiopi compliance agent “on a leash” through traceable tool calls, and forcing an insurance coverage check into a strict JSON schema instead of a prose summary. Three transferable lessons for any owner or CIO putting AI agents into production.

Read →