With 7 years of experience as a trainer in regulated industries, I have established a strong track record of deploying AI solutions across 8 EU countries. For the past 3 years, I have been actively working as a freelance consultant and trainer, specializing in the field of Artificial Intelligence.
I am a published author on AI-related topics and have been closely following the development of the AI Act since 2019.
Currently, I am actively collaborating with publicly traded companies, providing consulting and training services to ensure their compliance with the AI Act.
For referrals and further insights into my work, please visit my LinkedIn profile.
AI Act Trainings
As of March 13th, 2024, the European Union has ushered in a new era with the AI Act - the first comprehensive legislation governing the use of Artificial Intelligence systems. This game-changing regulation impacts every public authority and private organization operating within the EU Member States.
The clock is ticking. You have mere 6 months to ensure your AI systems are not on the prohibited list. If you're using high-risk AI, you'll need to comply with stringent requirements within 18-24 months. Failure to adhere to the AI Act's rules could result in more than just regulatory non-compliance – substantial fines and reputational damage loom for those who fall short.
The AI Act's broad definition of "AI system" casts a wide net, potentially including everything from machine learning models to advanced analytical tools and statistical data models. Determining which falls under the regulation's scope presents a significant challenge to achieve full compliance. However, the regulation framework demands more than just understanding its rules – it requires organizations to possess the technical skill to develop robust AI systems that are fully compliant, which necessitates having a team of highly skilled technical professionals.
To guide you through the regulatory framework, our services include extensive courses and specialized consultancy tailored specifically to your needs. We aim to equip your organization with both the knowledge and technical expertise required to craft AI solutions that not only comply with the AI Act's requirements but also excel in ethical standards.
Regulatory Awareness
Elevate AI Act awareness across your organization with our dynamic seminars and workshops. Each session is meticulously tailored to engage and educate various stakeholders, ensuring a comprehensive understanding of the regulation's impact and scope.
Cross-Functional Collaboration Solutions
We are bridging the gap between legal, IT, and business departments through our proven methodologies that foster effective collaboration, communication, and shared understanding – essential for achieving seamless AI Act compliance.
Specialized Courses
We offer specialized training sessions crafted for key roles including legal teams, auditors, and developers. These programs are designed to arm each group with the necessary knowledge and skills to effectively navigate the AI Act within their unique operational contexts.
AI Systems Compliance Audit
Benefit from our exhaustive consulting and auditing services, where we conduct detailed evaluations of your AI systems on a case-by-case basis. Our goal is to ensure your organization's practices are in full alignment with the AI Act, guaranteeing compliance and mitigating risk.
Deployment Enablement Service
We support organizations on the brink of deploying AI systems sourced from external providers. We act as a vital bridge between your team and your AI providers, gathering all necessary information and documentation. This hands-on approach not only ensures your new AI system's deployment is fully compliant with the AI Act regulation but also streamlines the process, enhancing your operational efficiency and mitigating compliance risks.
Due to high demand, we currently offer our courses exclusively in online format, with limited customization options available. Volume discounts are available.
AI Act Preparedness for Business Professionals
For whom is this course?
Are you a business stakeholder or a domain expert, who is around use cases that use smart algorithms? For example:
-
Sending e-mail marketing campaigns which are tailored to the individuals, or groups of clients.
-
Calculating default, or risk scores for retail clients.
-
Preventing fraud, or money laundering via data mining.
-
Predicting stock market movements
If any of the prior rings a bell with respect to your daily work, then this course is relevant for you. The EU is bringing the AI Act to the scene. This new regulation is aimed at AI algorithms, which are nowadays part of a wide range of banking use cases and products. If you would like to imagine it intuitively, while GDPR regulates the data that we work with, the AI Act regulates the algorithms that we build on top of it.
What will this course provide?
-
Essential knowledge of the AI Act: Definition of AI, Risk-Based approach for classification, General rules, Forbidden practices, High-risk AI classification & requirements
-
Triggers and alerts: When and why should I reach out to colleagues from legal?
-
Contacts and processes: What should I do, and whom exactly should I reach out to? What can I expect to happen afterwards?
Course delivery:
-
(2-)3 hours length
-
Online
-
Large classroom size (up to 30 participants)
Prerequisites:
This course has no prerequisites.
AI Act Preparedness for Developers
For whom is this course?
The EU AI Act is bringing forward many requirements for compliance of our AI systems. The important thing to realise is that a broad range of systems, that we usually would not recognize as AI, will now be defined as AI and hence regulated by the AI Act. Here are some examples:
-
Machine learning model that is predicting which product a customer will buy.
-
Logistic regression model calculating risk of default on loans.
-
Unsupervised machine learning model, acting as fraud detection engine.
-
Generative model used for generating internal documents, or code snippets.
-
Chatbot that customers might use to get answers to their most common questions.
If you are working on live systems such as the ones listed above, then this training is relevant for you, as you could consider yourself a developer of an AI system.
What will this course provide?
As a developer, what can you expect to happen in the upcoming months, or years, with respect to an AI Act?
-
Firstly, It will not be your task to study the details of this lengthy regulation, attempting to make your AI system compliant with it. It however can be expected that an internal legal representative will come to you, asking for information about the AI use case that you are responsible for, or are working on. It will be their task to make sure the use case is compliant and without your help and involvement, they would have a hard time in doing so. They will need you to provide them with as much objective information on the use case as possible. From this objective information, they will make the necessary judgments and recommendations for AI Act compliance.
-
Secondly, the AI Act is bringing forward practices that were not commonly adopted among developers of AI Systems. For instance, if an AI system is deemed as high-risk, in order to be compliant it has to have implemented components for protecting it against data poisoning, adversarial attacks, or it shall have thorough logging mechanisms with respect to transparency.
This course is aimed at these two points. After the training:
-
you will be able to effectively communicate with your internal legal colleagues, in order to make your AI use case compliant,
-
you will be aware of the new technical requirements that are brought forward by the AI Act.
Course content
-
Introduction to the EU AI Act: Basic terminology for efficient communication with legal people. Definition of AI, Risk-Based approach for classification, General rules, Forbidden practices, High-risk AI classification & requirements.
-
AI Systems Architecture: Tasks and approaches in AI modelling, machine learning model families, data engineering components, model deployment and infrastructure components.
-
AI Systems Tools & Frameworks: Interactive discussion with participants, Overview of common tools and their usefulness.
-
Writing (High-Risk) AI Technical Documentation: General description of the system, System development details, Monitoring, Functioning and control of the system
-
Developing Record Keeping for (High-Risk) AI System: Tools and frameworks for efficient logging, Understanding required and relevant logs, Data logging and provenance.
-
Developing Accurate, Robust and Secure Systems: Developing proper accuracy measures and benchmarks, Data and model centric robustness, Data integrity, Data and model poisoning, Model testing
Course delivery:
-
3 days length
-
Online
-
Partially hands-on (for relevant components)
-
Medium classroom size (10 participants). Such small group size allows for the part “AI Systems Tools & Frameworks” to be interactive. Participants inform each other on infrastructure and models they use. Participants hence get a broader perspective and feedback from the instructor.
-
Graduation examination assignment (up to 3 hours). i.e. small simulation of the expectations from a developer: report objectively on your use case, or write a technical documentation
Prerequisites:
-
Prior programming experience (e.g. R or Python)
-
Prior data science & machine learning experience (e.g. risk or propensity modelling)
Let's get in touch
Don't navigate the complex regulatory landscape alone. Reach out to me, and let's discuss how my consulting and training services can empower your organization to thrive in this new era of AI governance.