Team – March 27, 2018

Meet the Team – Ondrej

Welcome to our "Meet the Team" series where we profile the folks behind the Angee solution.


Name

Ondřej Plátek


Nationality

Czech


Position at Angee

Voice Lead


What you like about Angee

First, I am doing the job I’ve always wanted to do. I get to build a voice interface which is used by real people and learn about technology, people and how to communicate with them. Secondly, I find it deeply satisfying to cooperate with my colleagues and get things done. And lastly, I feel responsible for creating the voice solution, which is stressful sometimes, but rewarding. Also, I like that Angee is always improving on all levels. The software is obviously improving fast, but the team’s skills are also getting better by the day. The loop closes when the team cooperates and I can learn from my colleagues. It doesn’t matter whether I am learning new hacks in Vim, how to use Terraform & AWS, more about text-to-speech or am finally being introduced to embedded design. It just feels good to learn and it feels better to use the knowledge I’ve gained.

What/Who inspires you to do a great job

The improvement of the technology itself motivates me. I love to understand crucial algorithms and new discoveries. Lately, it has become quite easy to follow machine learning breakthroughs thanks to the open archive of research articles on arxiv.org. In fact, I follow only the most relevant papers thanks to http://www.arxiv-sanity.com/ because there are so many articles released these days. There are also a lot of open-source implementations to learn from. Despite quite a lot of low-level technical work at Angee, I try to follow papers around task-oriented dialogue systems, ASR, and related fields. At Angee, I enjoy regularly evaluating what is feasible to implement and what is not. It is definitely worth following guys from Cambridge, for example, Tsung-Hsien and Microsoft Research where I personally enjoy reading articles from Jason Williams. I hope we will have more time to implement principal approaches to task-oriented dialogue systems once we have a stable voice service. Personally, I would like to integrate some image recognition models, which I’ve played in my free time, in order to better distinguish people from pets and preventing false alarms for Angee.

Tell us what you do outside of work

I like to be outdoors and I try to spend as much time as possible with my family. I also try to spend most of my free time climbing with my wife and our friends who also have children. In the winter, I love skiing above all else.

 


Favorite songs

  • Marilyn Manson – Tainted Love

 

  • Black Label Society – Stillborn

 

We are changing the world. Do you want to join Ondrej? Contact patricia@meetangee.com

Angee

Angee is almost here - reserve yours now!

Reserve