Courses and events
AI training without the hype and without needless jargon
Practical programmes for teams and organisations that have to make decisions about artificial intelligence — plus an annual conference bringing research, industry, government and civil society to the same table.
What we are after
The most expensive decisions about artificial intelligence are not made by specialists. They are made by people who must decide without having time to become specialists. Our programmes are for exactly those people.
We do not teach fashionable tools and we do not promise transformation. We teach how these systems work, what they can and cannot do, how to check whether they work, what legal obligations they bring, and what an informed decision looks like inside a real organisation.
Every programme is built on the participants' own cases. A course in which a team works on its own data and its own processes is still useful six months later.
Programme formats
Four formats, from an orientation session for leadership to an applied programme running over several weeks.
Orientation session for leadership
- Who it is for
- Executives, boards, public decision-makers
- Duration
- half a day
What artificial intelligence is and is not, where it creates real value in an organisation like yours, what risks come with it, what legal obligations arise, and what to ask a supplier before signing.
By the end
- A checklist of questions for any proposed AI project
- A clear grasp of the difference between automation, analytics and generative AI
- Criteria for deciding what is and is not worth automating
Applied course for operational teams
- Who it is for
- Technical teams, analysts, product teams
- Duration
- 2–5 days, or modular across several weeks
A hands-on programme on the team's real cases: from framing the problem and preparing the data through to evaluating results and putting them into use. Work is done on the organisation's data and processes, not on textbook examples.
By the end
- One of the organisation's use cases taken from idea to evaluated prototype
- An evaluation method the team can repeat on its own
- Documentation of the technical decisions made along the way
Responsible AI and compliance
- Who it is for
- Legal, compliance, data protection, public procurement
- Duration
- 1–2 days
The regulatory framework applying to AI systems, the obligations that follow from it, how compliance is documented, and how legal requirements translate into verifiable technical requirements.
By the end
- Classification of your own systems by risk level
- A template for the technical documentation of an AI system
- Compliance criteria for tender specifications
Public or university workshop
- Who it is for
- Students, researchers, anyone interested
- Duration
- 2–4 hours
Open sessions on a specific topic: how large language models work, how an automated decision system is evaluated, what explainability means in practice, where current systems fail and why.
By the end
- Understanding the mechanism, not just the output
- Open materials that stay with participants
- A question session with no limits on subject
Annual conference
AIDCIA Forum — a day on artificial intelligence applied responsibly
Each year we run a conference on applying artificial intelligence responsibly, bringing research, the private sector, public institutions and civil society to the same table. The aim is not to present what is fashionable, but what works, what does not, and what we learned from both.
The first edition is in preparation. Date, venue and programme will be published here as soon as they are confirmed — and newsletter subscribers hear first.
Results, not promises
Talks grounded in work actually done, with the method stated and results that can be checked — including the ones that did not turn out as expected.
Four perspectives, one room
Researchers, industry practitioners, public institutions and civil society organisations on the same panels, not in separate rooms.
Working sessions
Parallel workshops where participants work on a case rather than just listen. Places are limited so that discussion stays possible.
Open materials
Talks and materials are published afterwards, so they are usable by those who could not attend.
Topics we cover
Each programme is assembled from these modules, according to the level and needs of the group.
Foundations
- How a model learns from data
- Classical machine learning and deep learning
- Large language models: what they actually do
- The current limits of these systems
Putting it to work
- Identifying use cases worth pursuing
- Data preparation and data quality
- Evaluating a model beyond accuracy
- Deployment and monitoring
Responsibility
- Bias, fairness and unequal effects
- Explainability and traceability
- Effective human oversight
- Documenting a system's decisions
Legal and organisational
- Data protection obligations
- Classification by risk level
- Requirements for procurement and suppliers
- Internal governance of AI systems
Frequently asked questions
Are the courses open, or only for organisations?
Both. Team programmes are arranged at an organisation's request and built around its cases. Public workshops and the annual conference are open to anyone interested.
Is technical knowledge required?
It depends on the format. The leadership session and the compliance course assume no programming. The applied course for operational teams assumes familiarity with data and usually with a programming language.
Can it be run at our offices?
Yes. Programmes can run on site, online or in a mixed format. We agree the format together, based on the team.
What language is it in?
Romanian or English, as the group prefers. Materials can be provided in both.
Is there a certificate?
Participants receive a certificate of attendance issued by the association. It is not an officially recognised qualification and we do not present it as one — it confirms that the programme was completed.
When is the annual conference?
The first edition is in preparation. We publish the date and programme on this page as soon as they are confirmed. Newsletter subscribers are told first.
Want a programme for your team?
Tell us who the participants are and what they should be able to do by the end. We build the programme from that answer, not from a standard offer.
Request a programmeFor announcements about public workshops and the annual conference, subscribe to the newsletter.