AI Labs profile: Rodrigo Pessoa

AI is changing how we think about and build learning products. Learnosity’s AI Labs is the team tasked with rising to the challenge. Here, software engineer Rodrigo Pessoa tells us about the motivations driving the innovations.

What do you do at Learnosity?

I‘m a software developer with the new AI Labs team. I’m responsible for developing state-of-the-art systems that leverage machine learning to produce innovative and intelligent solutions. My team and I are creating products using large language models to tackle important issues in education, such as question debiasing and improving accessibility.

How did you first get into AI? What was it that most appealed to you?

I truly got into AI when I started working as a research assistant in Trinity College Dublin after completing my Master in Science course. Back then, convolutional neural networks were the state-of-the-art and I worked on creating a model that was able to identify the video codec and container of a video file that could have obfuscated headers and even corrupted data. It turned out to be very performant and an interesting primer into the world of AI.

"There is a very strong culture of collaboration, which lets us move fast with product prototyping and development." Share on X

My educational background in signal processing, machine learning, and related areas already made me interested in AI as a topic. What appeals to me the most in AI, though, is the constant emergence of new behaviours and characteristics.

Perhaps this is most easily understood with large language models: only a short while ago, they were struggling to produce meaningful answers until models became large enough and started to produce great answers, to think step by step, to exhibit human-like behaviour (such as self-awareness) and get closer to passing a Turing test.

The AI Labs team is leading an exciting new era of innovation for Learnosity, and things are moving fast. What’s your experience been like so far?

Extremely positive. Everyone is eager to produce the latest and greatest products, which will move AI from its “shiny new toy” position into a powerful, paradigm-changing solution. 

The excitement, enthusiasm, and positivity is not only echoed by the team, but throughout the whole company. There is a very strong culture of collaboration, which lets us move fast with product prototyping and development.

I believe that knowing we are leveraging AI to help people, especially in the education sector, fuels our motivation.

Helping Learnosity become a leader in AI for assessment means making a very positive impact in education and, having come from an academic space, I know that tools that can aid productivity and deliver high-quality content are necessary now more than ever. Compared to many other industries, I feel that education has lagged behind in terms of significant breakthroughs and innovation, and I believe that AI will be pivotal to change that and deliver positive change.

The team has another new product on the way. Can you tell us a little about what it is and why it’s exciting?

Sure thing! Item Bank Health Check is a web application that lets users quickly perform myriad tasks regarding the overall ‘health’ of their item collections—such as identifying and fixing biases, adding missing alt texts to images, checking items’ WCAG compliance, transcribing audio and video to make content available in writing, and much more!

The product will help teachers, educators, and assessment agencies to deliver quality content at scale, ensuring their items adhere to their bias guidelines, and improve overall accessibility. It will also be a significant time saver by making routine checks and fixes automated, allowing for more focus and time spent on creative and pedagogical tasks. Students and assessment takers will benefit greatly from fairer, more accessible assessment.

Give us one big change you believe AI will have on assessment in the near-ish future?

I believe AI will have an impact on personalized and adaptive learning. 

As most of us know, AI has the ability to analyze vast amounts of data in real-time, which can significantly enhance assessments by tailoring and customizing them to each student’s learning path, level and capabilities. I believe AI will enable dynamic adjustments while a student is learning. If they answer a question correctly, the system could present a more challenging question next, and if they struggle, it could offer additional learning materials or simpler questions.

"The product will help teachers, educators, and assessment agencies to deliver quality content at scale, ensuring their items adhere to their bias guidelines, and improve overall accessibility." Share on X

I will take this a step further and say that I see great potential for learning in pairing AI, gamification, and real-time analytics. 

For example, imagine this: AI acts like a coordinator and tutor, guiding the students through the different syllabuses at each learner’s pace and level. Assessments become achievements, accomplishments, and challenges, which make them more engaging and less intimidating, reducing stress and test anxiety for the learner. Meanwhile, data collected by this system would be used to generate personalized reports for teachers and educators to gain more nuanced insights into a student’s strengths and weaknesses, allowing more targeted interventions and personalized support.