Using ESCO for training synthetic digital personas in AI-driven research programmes

Introduction

This testimonial discusses how Digividuals AI R+D uses ESCO as a core taxonomy for defining the business layer of our Synthetic Digital Personas - Digividuals, to facilitate AI-driven research and building Business Support Systems (BSS).

Our Organisation

Advanced Analytica AI R+D is a cutting-edge research and product firm founded in 2013, dedicated to revolutionising market research and business support using AI. We create Digividuals that provide instant, accurate, and comprehensive business insights. Our proprietary AI Character Engine (AICE) leverages the ESCO framework to develop intelligent business assistants tailored to our clients’ unique needs. We serve businesses across all industries, helping them streamline processes, enhance decision-making, and drive success through AI-driven research solutions.

How we leverage ESCO in our services

Advanced Analytica AI R+D has integrated the ESCO framework as the core taxonomy for building our portfolio of Digividuals AI for the EU labour market. This enables us to provide B2B research and product services and deploy AI-driven business support solutions. We use the ESCO taxonomy to dynamically map Digividuals business roles, skills, and competencies to standardise the skills and competencies layer of Digividuals. It forms the basis of AICE, enabling precise and relevant interactions, decision-making processes, and insights. Our system supports clients in understanding complex business scenarios and optimising strategies, ensuring effective and personalised AI-driven solutions.

3. What were some of the challenges you encountered in implementing ESCO?

Implementing ESCO within AICE presented several challenges. Firstly, ESCO’s extensive taxonomy for EU occupations’ skill and competency requirements made it complex to differentiate between ‘essential’ and ‘optional’ skills for specific business roles, making it difficult to prioritise and customise these for our Digividuals. Secondly, translating the vast array of professions and skill requirements into training data for Digividuals was challenging, ensuring each Digividual accurately mirrored real-world roles. Lastly, seamlessly integrating ESCO’s detailed taxonomy with our AICE required meticulous mapping and continuous updates to maintain relevance and accuracy.

4. What is the key ingredient in ESCO that made you take the decision to use it in your system?

The key ingredient in ESCO that led us to integrate it into the AICE was its ability to provide a standardised and comprehensive model for training Digividuals with skills and competencies across diverse occupations. ESCO’s detailed taxonomy allows us to accurately train Digividuals with business roles and skills layers, creating precise business knowledge for our Digividuals. This ensures our AI-driven Digividuals are based on a robust and managed dataset, allowing us to build relevant, reliable, and tailored solutions to meet the specific needs of our clients, facilitating effective decision-making and strategic planning.

5. What was the level of acceptance by your partners/other national stakeholders?

Despite initial unfamiliarity among our partners and customers with ESCO, its integration into AICE has received widespread acceptance and overwhelmingly positive feedback. We have collaborated closely with clients who were initially unaware of ESCO and the value it brings to B2B research projects and developing highly intelligent Digividuals. This collaboration has been particularly fruitful in enhancing the effectiveness and adoption of our AI-driven research services and business support systems.

6. What are some of your recommendations for other stakeholders looking to implement ESCO in their system?

For stakeholders looking to implement ESCO in their AI systems, it’s crucial to validate the relevance of skills and competencies through prototyping and test environments. Familiarise yourself with the ESCO API and classification, including its hierarchy and the relationships between skills, competences, qualifications, and occupations, and tailor it to your specific persona context. Ensuring stakeholder alignment is vital for adoption. Additionally, regularly updating the ESCO data to maintain accuracy and relevance is essential. This approach helps leverage ESCO’s comprehensive framework for improved accuracy and efficiency in AI-driven solutions.

7. How has ESCO helped your business/organisation? What are its advantages and disadvantages from your standpoint?

ESCO has been instrumental in standardising the job roles and skills layers for building Digividuals. It serves as a core data source, enabling us to derive meaningful insights and integrate these into AICE. This framework has become a fundamental element of our business strategy, enhancing the accuracy and relevance of our Digividuals. By providing greater clarity about the capabilities of business agents, ESCO fosters trust and reliability among our clients. However, the extensive data and complexity of differentiating between essential and optional skills can pose challenges.