DevCon 2019: AI & Machine Learning Interview & bits and pieces from our last edition

3 September 2020 news

For 2020 and this November, we raised the bar and created a different concept event that showcases 4 weeks full of content on Java, Cybersecurity, JavaScript and Web & Mobile, with 30-minute keynotes, workshops with 25-30 participants, 1.30 h masterclasses and many more suprises.

Now back to our first edition and one insightful interview

At our first Edition of DevCon in 2019, we deep dived into the most discussed topics and technologies of the moment with 4 dedicated conferences on AI & Machine Learning, Security, QA & Testing, and Java.

DevCon 2019 was a premium event where senior IT professionals gathered for 3 days to share their best inputs and experiences and connect to the latest trends in technology and contribute to the future of the tech industry. The event emphasized the most important tools, key learnings & resources, or advanced tips & tricks from powerful top-notch local and international speakers.

Celebrating the success of our first day and marking the AI & Machine Learning Conference, we had the pleasure to steal some precious info from Abraham Kang (CTO at GEEE), speaker extraordinaire and a guy so charismatic that made our day way better.

His Talk – MLSec and how your ML applications may be at risk – offered a deeper understanding of machine learning adversarial examples. Similar to fuzzing, adversarial samples/examples exploit weaknesses in how input is processed. Adversarial Examples are an inherent weakness in almost every machine learning model.

If you wanna find out more and expand your know-how about machine learning, keep reading the following interview, Abraham has some wise words towards AI & machine learning and how deep learning continues to impact our lives.


Abraham Kang is fascinated with the nuanced details associated with machine learning algorithms, programming languages and their associated APIs and he’s currently doing research in security for machine learning. He is focused on machine learning/AI, web, application frameworks, blockchain smart contracts, intelligent assistants, and mobile security and has presented his findings at RSA USA, Black Hat USA, DEFCON, OWASP AppSec USA, and BSIDES.


Hi, Abraham! What did you like most about the first edition of DevCon & how did you feel the vibe of the conference?

Abraham Kang: I really liked the diversity of topics that DevCon provides in tech, it expands a lot of areas and it’s really relevant to today’s environment and a great resource.

I also liked the interaction with people, usually when I give this talk, depending on where it is, sometimes the response would be a little bit less and here at DevCon, there was a lot of interaction. The people are very smart in Romania and I’m very impressed with the talent that is here and the intelligence that it is within the room.

What did you like most about the Romanian audience of IT professionals?

Abraham Kang: Romanians are very sharp, they understand a lot of the concepts very quickly and they seem to have a great understanding of machine learning and the foundation and the knowledge to understand the adversarial tax and the machine learning security vulnerabilities that we discussed.

What was your keynote about?

Abraham Kang: I came here to make people aware of the machine learning security vulnerabilities and to help them understand how they work and how to defend themselves against them and what they have to look out for.

Do you have a message for the local IT community in Romania?

Abraham Kang: The message that I would have is basically keep on doing well on machine learning and keep on expanding cause that’s the way things are going and you have a great population & a great set of people and I think the future is very bright for Romania.

What was the main idea of your talk?

Abraham Kang: To basically understand the tax that it can occur on machine learning models and how to defend them.

What do you think about the Romanian tech scene?

Abraham Kang: Very cool scene, I was talking to different people before the talk and there’s a diversity of companies here with a lot of different tech capabilities that it’s really great to see.

How do you predict the future of what you showed us?

Abraham Kang: A lot of what I’ve showed you with deep learning is very good for certain things like pattern & identifying patterns & large data but the future is going to be something different, to get machine learning models that actually think and discern, in generalised world, I think we’re gonna need a different model & that’s where I’m hoping that Romanians with their awesome intelligence and understanding will come up with the next big thing in machine learning.


Stay tuned and check more on

Book your seat below and join our journey:

?‍? Java Week, 3-7 November
?‍? JavaScript Week, 10-14 November
?‍? Web&Mobile Week, 17-21 November
?‍? Cybersecurity Week, 24-28 November.