Cookie settings

Some cookies are needed for the site to work. Google Analytics starts only if you agree. Until then, it sends nothing to Google.

Supports security, sign-in, forms, language, and saving your choice.

Shows us which pages people read. It starts only if you agree.

Shows us whether our Facebook and Instagram ads bring visits and sign-ups (Meta Pixel). It starts only if you agree.

Read the cookie policy
Skip to content

ABOUT AI FOCUS

Built from practice, not presentation.

AI Focus is the AI Architecture & Enablement practice. We build courses on your team's work, find where AI makes sense, implement the systems and keep them running.

How we work

Sometimes the answer is no AI

We look at how the team works before we recommend anything. If a spreadsheet, a better process or postponing the project would work better than AI, we say so openly. We are paid for judgement, not enthusiasm.

  • We start with a conversation

    About teams, roles and what exactly needs to change. The answer can be a course, a project or neither.

  • Offer

    In writing: what we deliver, what is not included, who does what and the price. The offer fixes the scope, the capacity and the responsibilities.

  • NDA on request

    We sign our standard non-disclosure agreement (NDA) before the first conversation if you ask for one. A company's own agreement is negotiated and can change the price.

  • The tools

    We recommend Claude, with Claude licences for the people taking part. We also work with ChatGPT if your team already uses it. The offer says which licences are needed.

The principles we work by

The aim is not tool adoption for its own sake, but useful, governed capability that fits how the organization works.

  1. Work before tools

    Start with the work: the task, information, decisions, expertise, bottlenecks, and repeated effort. Only then decide whether AI belongs in the process.

  2. Capability before productivity

    Durable capability means people can decide when to use AI, evaluate its output, improve it, and adapt as tools and models change.

  3. Knowledge before automation

    Automation built on unclear or stale knowledge produces faster mistakes.

  4. People remain responsible

    AI can assist with bounded work. People remain responsible for purpose, judgment, approval, accountability, ethics, prioritization, communication, and leadership.

  5. Systems over isolated tools

    Architecture connects AI to the people, knowledge, systems, and defined work that make it useful.

  6. Adoption is continuous

    Useful capability needs learning, maintenance, governance, review, and improvement over time.

Who teaches and builds

The written offer says who does what, on our side and on yours.

Alexandru Mareș

Alexandru Mareș

AI systems architect · strategy and delivery

Builds AI systems for companies, from architecture to production.

Claudiu Rentea

Claudiu Rentea

AI expert · design, UI/UX and marketing

Brings AI, design and marketing together to build digital products.

Tell us what your team works on

We reply with the right step, even when that step is a scheduled course and not a project.

The first conversation is free. We reply within 3 working days. On request, we sign our standard confidentiality agreement (NDA) before the conversation.

Or write directly to office@aifocus.ro