Searching for a job online can be a long, frustrating and inefficient process. Hours spent reviewing job offers, reading requirements that don’t match your profile or sending CVs to positions that aren’t really right for you. With this in mind, three students from Somorrostro’s postgraduate course in Artificial Intelligence and Big Data (IA4) —Oier Cadierno, Jon Medina and Samuel Rivera— have developed Job Matcher AI, a project that applies artificial intelligence to optimise job searches and improve the match between candidates and job offers.
This project was developed as part of one of the course challenges and is a clear example of how the centre’s AI training is geared towards solving real problems, combining technology, data analysis and ethical reflection.
An AI that understands your profile… and the offers
Job Matcher AI is an intelligent chatbot that matches the characteristics of a CV with job offers posted on the specialised website Manfred, a portal focused on technology profiles.
The system analyses technical skills, experience, desired salary and location, and returns the offers that best match the user’s profile, avoiding wasted time and endless searches.
‘The idea is simple: you don’t have to go through each offer to see if it’s a good fit. The AI does that work for you,’ explain the students.
Why Manfred? The importance of good data
One of the key aspects of the project has been the choice of data source. The students chose Manfred over other better-known portals such as InfoJobs or Indeed for one very clear reason: the quality and structure of the information.
Manfred presents job offers with clear and consistent data: salary, type of work (remote or on-site), required experience and technical skills. This greatly facilitates the work of artificial intelligence.
‘In AI, there is a basic rule: if you feed a system with bad data, the result will also be bad. Manfred allowed us to work with well-presented and reliable information,’ they point out.
How Job Matcher AI works
:
- Extracción de datos
El sistema recoge automáticamente la información pública de las ofertas de empleo de Manfred y la almacena en una base de datos. - Procesamiento inteligente
Esa información se organiza y se prepara para que pueda ser interpretada por la IA. El sistema “entiende” qué pide cada oferta y cómo se relaciona con las habilidades del usuario. - Conversación con el usuario
A través de un chatbot, el usuario introduce su perfil o incluso su currículum, y la IA compara esos datos con las ofertas disponibles para mostrar las que realmente encajan.
“Es como darle al sistema un montón de ofertas para que se las aprenda y luego preguntarle: ¿cuáles son las mejores para mí?”, explican Oier, Jon y Samuel durante la entrevista
Time-saving and more efficient searches
One of the great advantages of the project is the time it saves. Many people spend hours reviewing job offers that they ultimately discard because they do not meet the minimum requirements.
‘With Job Matcher AI, you avoid reading offers that require five years of experience when you only have one. The AI filters them out before you waste your time,’ they say.
The system also allows you to prioritise aspects such as salary, location or remote working, offering a much more personalised experience.
A project with an ethical and legal perspective
Beyond the technical aspects, the project has served to reflect on the ethical and legal implications of using data in artificial intelligence.
During development, debates arose about data protection, corporate consent and the responsible use of public information. Although the project works only with publicly accessible data, the students emphasise the importance of establishing clear limits and governance in this type of system.
‘The technology can be developed, but it cannot always be applied directly. Legislation and data protection must be taken into account,’ they explain.
Learning AI by solving real-world problems
Job Matcher AI was developed in just three weeks, with more than 200 hours of work within the framework of the postgraduate programme, and is not intended as a commercial product, but rather as an advanced learning project.
However, its potential is evident. The system could be expanded to other job portals, incorporate new functionalities or be integrated into career guidance platforms.
‘These types of challenges allow us to apply everything we learn: databases, language models, analysis of alternatives, ethics, and system maintenance,’ they emphasise.
Training that connects with the professional world
This project reflects Somorrostro’s approach to Artificial Intelligence and Big Data: to train professionals who are able to understand technology, apply it critically and adapt to a constantly changing work environment.
Job Matcher AI not only demonstrates the technical level of the IA4 postgraduate students, but also their ability to identify real needs and propose intelligent solutions.
Una vez más, la IA deja de ser algo abstracto para convertirse en una herramienta útil, cercana y con impacto directo en la vida de las personas.
