Bruno Fernandes

Every single business can take advantage of AI

March 3, 2024

You are an AI professor at Minho University in Portugal and the founder and CEO of PluggableAI. Could you tell me a bit more about how you ended up here and when technology and AI were part of you?

Well, I’m from a small village in the north of Portugal. A village where not much happens. But still, opportunities are there for everyone! In 2009 I went to the University of Minho to study Software Engineering and, from that moment onwards, I’ve been living in Braga. After my master's, I decided to get my PhD in Informatics (AI specialized) with a particular goal – the goal was to have time to conduct research in domains that could allow me to, later, launch a startup later. That, indeed, happened. A few years later, I launched PluggableAI, together with Ana Bela, my co-founder, which holds a PhD in Neurosciences. 

What inspired you to found PluggableAI?

Since I remember, I always had the goal of (at least) trying to launch my own business. That’s what I’m trying now, with Pluggable, but I can tell you that it is much harder than I was expecting! And, unfortunately, Portugal is not kind to small businesses…

Your product is focused on Analytics and smart notification systems. Could you explain them a bit more?

Pluggable works on the concept of Meaningful Mobile Engagement. By using disruptive AI models and non-intrusive mobile sensors, we created a Smart Notification System, i.e., an intelligent system for creating, optimizing, and delivering push notifications at the best time to every single person, allowing companies to communicate efficiently and effectively (personalized communication), thus increasing user engagement, app retention, brand perception, and awareness, while eliminating feelings of spam and intrusion. Pluggable’s solutions can be incorporated, by any mobile app, in their source code, adding a layer of intelligence to any mobile app. In summary, Pluggable’s solutions use disruptive rule engines, Deep Learning models, and more than 60 non-intrusive mobile sensors (such as the accelerometer, the gyroscope, the temperature of the CPU, …) for on-device inference (all models run in the customer's mobile in less than 100 milliseconds). It is worth noting that Pluggable solutions are entirely non-intrusive, fully respecting the user’s privacy.

Pluggable is funded by Portugal Ventures as a pre-seed funding. As a founder, you were part of this process. How challenging it is to get funded and is the plan to get seed funding in the future?

We are now preparing a seed round that is likely to happen in the next weeks. Still, even though I find it easy to have people interested in investing in Pluggable, the process of negotiating the term sheet and the investment agreements is a long-running process, that can take many weeks. I heard, recently, a sentence that I really appreciated: “Fundraising is a full-time job!” – well, I now tend to agree!

How possible or impossible is it to bootstrap your AI product nowadays?

That is always possible! Still, there are risks that you must consider and that may affect your future. If any of you are in that position, I would advise you to apply to pre-acceleration programs (Startup Bragas has very good ones) and take advantage of the networking and skills you will learn in such programs! That’s how we started Pluggable.

What is the goal for 2024 for Pluggable AI?

Our motto for 2024 is to become the biggest Portuguese company in digital communication. It’s hard, nearly impossible to accomplish, but we will fight for it!

How do you see AI impacting small businesses in the coming years?

We can already see the impact of AI technology such as ChatGPT Gemini, Midjourney… My opinion is that the future is promising, mainly for those who are focused on “applying” instead of “creating”.

As an AI professor, what key aspects of machine learning and autonomous systems do you emphasize to your students?

It is not easy to put to give you an answer in a few lines. Still, some of my mine main goals are to :

  1. Ensure that the students, besides knowing how to apply the technology, know what they are applying, i.e., that they know the theory behind the models they are building;
  2. To have students solving real-life problems instead of dummy ones (we create intelligent bots, solve traffic flow forecasting issues, we create technology for the elderly, …);
  3. To have a healthy competition between students – every year we launch a new edition of a ML competition in Kaggle were the students compete to develop the best possible solution to a real-life problem (this year the goal was to develop the best model to predict the energy produced by solar panels considering several variables such as the need for energy, the weather conditions, the location of the panels, …).

Can you share one of the most exciting or challenging research projects you've supervised?

Well, the most challenging is easy – a project where we were developing, entirely from scratch, a tokenized blockchain to store open data (called Knowledge). It was a very interesting project that allowed us to make important findings but, unfortunately, still required huge amounts of work to see the daylight. Interesting projects, I have many: the ability to detect boredom in smartphones; applying non-intrusive sensorization for fan engagement; a sensor that we conceived that was able to infer the amount of people in a particular area without cameras or the need for people to do anything … and a few others!

How do you translate complex AI and machine learning concepts into practical solutions for example boredom detection or time series forecasting?

It is not easy to translate such complex knowledge and translate it here, in a few lines! Still, I believe that a strong practical approach is the best way to learn the theoretical components. I find it difficult to properly deliver “a message” just by speaking – it is easier to understand it when you put it into practice. 

What are some common challenges you face as an AI researcher, and how do you approach overcoming them?

In the past, the main issue was to have the proper machines to conceive the models that we aimed for – strong GPUs were needed, in good machines. Now, that is easier to overcome with cloud solutions – in our case, we use Azure and we are part of Microsoft Founders Hub. If you are a startup and need credits to develop your solution in the cloud, my advice would be to join Microsoft Founders Hub. For us, it was worth it (many, many credits and multiple platform subscriptions entirely free).

What advice do you have for small businesses looking to integrate AI into their operations?

For small businesses, if you have 0 knowledge of the matter, that would be extremely difficult. I guess it would depend on the ability to have someone who, if not knowledgeable in AI, at least is motivated to learn it!

As AI becomes more integrated into various aspects of society, how do you approach ethical considerations in your research and development work, especially in areas like affective computing?

As I’ve said, privacy and respect for the user’s data are a core values for Pluggable. That’s why we only use non-intrusive sensors and do not collect any data that refers to the user. We do our best to ensure full compliance with the user’s privacy and can only expect that other companies do the same.

Can small businesses leverage machine learning to gain a competitive edge, and what potential challenges should they be mindful of when implementing ML solutions?

That’s the issue – they can’t, at least the typical small Portuguese businesses can’t. We can ask a small retailer to implement ML solutions for stock forecasting, for example. And I don’t think that the small retailer will have the financial availability to pay for that, unfortunately, even though it could significantly improve the business. Again, the problem is not with the small retailer, but instead with the government and the lack of incentives (and huge taxes). 

I guess you can have some busy schedules with Pluggable and as a teacher. Also, you seem to like sports a lot. How do you manage to find the time? Any technique that just works for you?

Well, to be honest doing sports is what allows me to keep my head clear, my stress levels controlled, and to maintain a balance between working and resting. So, doing sports is mandatory for me – usually, late in the afternoon! It is non-negotiable!

What is the business area that going to change the more with AI?

Every single business can take advantage of AI. That is what I believe will change the most – any business, regardless of its core domain, will be able to take advantage of AI to develop its business. What we should guarantee is that everyone has the same opportunities.

Favorite tool and why?

For AI, TensorFlow. To develop code, Visual Studio Code. For CrossFit, thumb tape! :)

Word of advice

It is all about the journey! The outcome can be better or worse – it is the journey that makes it worth it!

One book, one person to follow/ that inspires you, and one music?

Book-The Count of Monte Cristo;

Person No one;

Music- Shine On You Crazy Diamond - Pink Floyd

Bio

Bruno Fernandes holds a Ph.D. in Computer Science, Artificial Intelligence specialized, from the University of Minho, having achieved the highest rating of Very Good with unanimous approval. He also holds a bachelor's and master's degree in Informatics Engineering, both obtained at the University of Minho. Since 2018, Bruno has been a lecturer at the Department of Informatics of the University of Minho, being responsible for lecturing courses in the field of Computer Science, particularly in the domains of artificial intelligence and data science, to programs such as the Master's in Informatics Engineering, the Master's in Mathematics and Computation, and the Master's in Law and Informatics. Until 2022, Bruno Fernandes conducted scientific research at the Synthetic Intelligence Laboratory within the ALGORTIMI Center of the University of Minho. He is also a researcher in the Artificial Intelligence and Health Research Unit at CESPU (Cooperative for Higher Education, Polytechnic, and University). Bruno is also the founder and CEO of PluggableAI, a leading Portuguese startup in the fields of non-intrusive sensorization and artificial intelligence in mobile devices.

Latest Posts