Curiosity and innovation are at the heart of Avature's Machine Learning team, where cutting-edge AI solutions are reshaping the future of HR. In this new edition of the #AvaturiansExperiences series, Luis Gascó, an NLP expert with a rich background in academia, shares how his transition to Avature has allowed him to combine scientific exploration with practical applications that drive real-world impact. From pioneering neural technologies for skills extraction to advancing Generative AI architecture, Luis and his team are continuously pushing boundaries. Dive into this inspiring story of passion, learning, and collaboration, and discover how AI is transforming talent management at Avature.
*
When and how did you meet Avature?
I discovered Avature by chance. Before joining the company, I was working as a research engineer in biomedical NLP (Natural Language Processing) at a Spanish research center. I’ve always been curious about how language technologies and machine learning can be applied across various fields, so after being able to apply them to urban management, environmental studies and healthcare, I wanted to explore their potential in HR. While taking a look at what companies in the sector were doing, I discovered Avature’s Machine Learning developments, which inspired me to work on a personal project, applying AI to labor data. A few months later, Avature recruiters got in touch with me, and after a few interviews, I officially became part of the team as Senior Machine Learning Engineer in February 2024. I think that my previous knowledge on existing technologies applied to the sector, and my insights about potential transfer of biomedical NLP developments into the HR field played a significant role in my selection.
Can you tell us about your transition from academic research to working in a tech company? What led you to make that shift?
The transition from academic research to working in a tech company was both exciting and natural for me. While there are clear differences between academia and the tech industry, I found several overlaps that made the shift smoother. For example, academic research develops strategic planning skills for long-term projects, where plans often need to be adapted to new technological advancements. This adaptability and pragmatism are key skills that align perfectly with what’s needed at a company like Avature, where we follow a roadmap for developing new features while being flexible enough to incorporate emerging technologies in the process.
For me, transitioning to the tech industry also meant acquiring additional skills, such as expertise in managing data, software engineering principles, and Machine Learning Operations (MLOps). Through this shift, I also learned the importance of collaborating with diverse internal areas, including Development, Engineering, Product Management, among many others.
What led me to make the shift was my desire to see a real-world impact of my work. In academia, the outcomes are often more abstract, while at Avature I can see how my contributions directly benefit customers, colleagues, and other stakeholders. At the same time, I’ve been able to continue making scientific contributions by attending research conferences and staying actively involved in that community. This mix of applied work and staying connected to academic research isn’t something you find in most industry groups working on machine learning, which I think is such a cool part of Avature’s culture.
With so many opportunities in AI today, what drew you to Avature’s Machine Learning team?
I was drawn to the team from the very beginning. It’s a unique crew with the perfect mix of highly experienced software developers, artificial intelligence experts, and Avaturians with research backgrounds. On top of that, being part of a relatively small but always-evolving team means there’s great potential for growth, which I believe will allow us to have an even greater impact in the coming years.
What project or development in the Machine Learning team are you currently most excited about?
I’m currently working in the application of AI in the skills area. After some months of hard-work, we are going to launch the first production-ready system for skills extraction and normalization using neural technologies, which represents a significant improvement over what’s currently available in the Avature platform. We’re also working on creating tools to accelerate the expansion of our skill dataset by leveraging semantic technologies. All these advancements form the foundation for further improving our technology, enabling us to deliver even better results in the features built on top of these capabilities.
As for the team’s efforts, I’m particularly excited about the progress done on developing the new Generative AI architecture. This will enable the inclusion of a wide range of impactful features on the platform and allow us to implement new GenAI features more quickly.
I heard you often participate in different conferences and talks about AI and NLP in HR, and in your most recent presentations, you discussed how NLP is shaping the future of HR. What do you believe are the most exciting ways AI can transform recruiting and talent management?
I absolutely love taking part in these kinds of events! They are a great opportunity to build a network around NLP applied to HR. Since I started working at Avature, I’ve had the opportunity to give these talks to university students and other professionals, which allow us to share our vision for AI in HR while also connecting with new talent eager to make an impact in this exciting and rapidly evolving field.
In these talks, I always highlight how the revolution driven by the development of Large Language Models (LLMs) is transforming talent management, enabling more dynamic and natural interactions with data, something impossible just a few years ago. At the same time, I usually highlight the importance of continuing to develop “traditional non-GenAI” specialized systems that can process massive amounts of data more efficiently and, in some cases, deliver even better results.
For example, if we want a system to create tailored job descriptions for a role like Data Engineer, we could fully rely on the knowledge of an LLM, but an even better approach might involve using traditional AI to identify and extract the most relevant skills and educational requirements from past data. This information can then be fed into the LLM, guiding it to produce a more accurate and targeted final result, ultimately blending the strengths of both approaches.
What’s your favorite part about engaging with academic audiences and representing Avature in university settings?
As a person with a professional background in academia, I’ve always enjoyed interacting with students and sharing my experience and knowledge with them. In fact, I’m also a professor of NLP in two university master’s programs, which gives me the opportunity to stay connected with the academic community and engage with new generations. Representing Avature in university settings allows me to show students how all the knowledge they are acquiring can be applied in practical scenarios showcasing real application examples, something that can sometimes be overlooked during their academic journey. It is incredibly rewarding to help them see the tangible value of their knowledge.
What contributions does Avature make to the AI community, whether through research or published papers?
Our goal is to make a significant impact within the AI community and position Avature as a leader in this field. To achieve this, we actively participate in scientific events and foster collaborations between universities and the industry.
In 2024, we’ve been especially active. Highlights include:
- Publishing advancements in Graph Neural Networks and chatbots to the field of Human Resources at SEPLN 2024, the Spanish Conference on Natural Language Processing.
- Having two works accepted at the 4th Workshop on Recommender Systems for Human Resources (RecSys in HR 2024), showcasing our progress in multilingual job title matching—covering both technical innovations and the introduction of new model evaluation benchmarks.
- Launching a new initiative called TalentCLEF, the first community evaluation campaign focused on NLP applications in Human Capital Management.
- Establishing the HiTZ Chair for Artificial Intelligence and Language Technology in collaboration with the University of the Basque Country. This partnership provides our team with a framework to collaborate with one of the most prominent universities in the field of language technologies.
And there’s more! My colleagues and I actively participate in other professional events, such as presenting at Tecnoling 2024 and giving talks in European projects like AI4Labour.
All these engagements allow us to stay connected with the broader research community while showcasing our work and fostering knowledge exchange.
AI is something very new for everyone and is evolving rapidly—what challenges does this bring for your team and work dynamic?
The AI field, and particularly NLP, has seen some massive changes in recent years, completely reshaping the way we have to work. We’ve gone from using statistical models to deep neural networks, which do a much better job at understanding the meaning of words. On top of that, a wave of disruptive technologies has come along, such as static embeddings or transformer architectures, boosting performance and opening up all kinds of new application possibilities.
To keep up in such a fast-paced environment, our team needs to stay curious and passionate about the technologies we work with. This means keeping up with the latest developments that could fit our use cases by diving into scientific papers from universities and big players. It also calls for a bit of creativity, such as taking breakthroughs in fields like biomedical NLP and finding ways to apply them to HR challenges.
How do you envision AI at Avature evolving in the next few years? What role would you like to play in shaping that future?
I believe that AI is going to have a cross-functional impact at Avature. It’s hard to predict the future, but I think our efforts to build the foundational AI blocks and architecture will allow us to add new features to the main app much faster. I’m confident this will have a big impact on both our current and future customers, but also in our internal processes, such as using tools like our Internal Knowledge Base (IKB) to make finding information much easier.
I joined the team some months ago, and I feel it’s been a great professional decision. I’m really comfortable with my daily tasks, as they allow me to combine my personal interests in science and technology with delivering outputs that have real effects. I’m also able to interact with product experts and shape the challenges they face. Personally, I’d like to keep contributing my NLP expertise to the skills area, enhancing the capabilities we currently offer. I’m also eager to be part of developing GenAI solutions in this space to create even greater value for our customers, while continuing to grow professionally within the company.
As someone who has built their career in AI and machine learning, what advice would you give to those with an academic background who might be considering a similar move into industry?
My advice is not to be afraid of transitioning to industry! In academia, it’s common to assume that corporate environments are fast-paced and don’t allow much space for creativity. Still, the rapid technological advancements in AI have led companies to value the skills acquired in academia as essential for driving innovation. This shift opens up opportunities to apply and evolve those abilities within a company, make a meaningful impact, and continue to grow creatively and professionally.