Developer stories — Ashwin Nanjappa from Singapore 🇸🇬
The idea of this series was to talk to a bunch of peers working in different environment and countries all over the world, so it was great to start from someone that can give us some insights about a tech hub of the likes Singapore.
Ashwin was happy to share a bit about his work and answer my questions via email — and even though I’ve never been involved in his particular field — AI and Deep Learning (and know very little around the subject) — his contagious enthusiasm made me curious to learn more.
Let’s get down to the questions:
Q. Where do you live, work, what’s your job title? Introduce yourself however you prefer.
I work at Visenze in Singapore. We are an AI startup, focused on applying computer vision to fashion e-commerce to create better shopping experiences and drive sales. Our solutions are used by some of the biggest names in fashion like ASOS (UK), Myntra (India) and Zalora (Asia).
I have a PhD in Computer Science from the National University of Singapore for designing and implementing massively-parallel GPU algorithms to solve the 3D Delaunay triangulation problem.
Prior to Visenze, I had worked as a post-doctoral research fellow at ASTAR Bioinformatics Institute in Singapore on designing and implementing vision algorithms for human hand pose estimation and mouse pose estimation.
At Visenze, I’m a senior research engineer in the Algorithms team. I work on designing and implementing computer vision pipelines for detection and search of fashion products in images taken by users on their smartphones or from product images.
My daily work spans from devising new computer vision solutions to automating data cleaning and annotation to pure software engineering problems and code optimisation.
We regularly study, apply and improve on the cutting-edge research in deep learning, object detection and visual search. My team is filled with the smartest PhDs in computer vision and it’s exciting to learn from them every day!
Q. What sparked your very first interest for development?
One of the first games that hooked me in high school in India was Prince of Persia. I’m a pretty pathetic gamer, but I was wowed by the graphics and gameplay.
It was an amazing new sensory experience built inside a computer in code. I was blown away that a single person could build such a system and I knew that I wanted to learn all about working with this wonderful machine and create virtual systems using it.
Back then, there were these books in the school library with Basic code for creating simple games. You could type out a few pages of code, fix the bugs and lo behold you had a simple game inviting you in to play it! It was (and still is) all pretty magical.
Q. What made you decide to make a career out of it, and how did you move towards it?
Back when I signed up for a Bachelor’s degree in Computer Science, there were not many jobs in the software industry. But by the time I finished the degree, software industry was booming. Back then, Indian software companies were focused on building enterprise solutions for companies in US, Europe and Japan.
However, I was lucky enough to work on middleware for DVD players created by Philips and Yamaha. Later, I worked on similar systems for Motorola. This was the pre-iPhone era and my code shipped in some of the most popular phones in the world back then. I later decided to get a PhD and transitioned towards computer vision.
Q. How did your environment contribute to your career?
My family had always been pretty supportive of my interests. I had always been a bit geeky, with zero interest in sports and a huge love for literature and science since school. The best birthday gift I got from my parents was a membership to the local library.
My middle and high school years ended up being a hazy blur of chomping through books as fast as I could! I read everything that was available from fiction to history (World War II!) to encyclopedias.
My parents encouraged me with whatever I wished for, like regularly signing up for chemistry and biology labs over summer to do some real lab work.
They were pretty patient while I broke up all the electronic devices at home and was usually unsuccessful in putting them back together! There was no one in my family or neighbourhood with these eclectic interests though and I was pretty alone at home in this path that life was leading me through.
However, there were a handful of like-minded geeks in my classes, so I did not feel that lonely about all this.
My dad has a fascination with literature and he writes too. I ended up being interested in writing too. Though my literary jaunts did not go far, I ended up fusing it with my love of technology and started blogging regularly at codeyarns.com.
Over the years, I have written close to 2000 posts and the blog has been surprisingly popular, seeing more than 4 million visits. In the past, I have also been an editor and writer for some online tech outlets, like the Linux Journal and tech magazines.
Q. Are you an immigrant? Would you relocate for work?
I moved from India to Singapore for my PhD and I stayed on after that for work here. Though it is tiny, Singapore has some great universities and a good ecosystem of startups in AI, fin tech and web commerce. I would relocate if any exciting opportunities piqued my interest. Though I must admit relocation is harder now that I have a loving family.
Q. In the unrealistic scenario in which the internet shut down all of the sudden, what else would you do?
If the shutdown is temporary, I would take the opportunity to travel. Experiencing new cultures has always been an interest of mine. If the shutdown is permanent, then we still have the fantastic world of embedded and electronic systems to tinker with and build.
I would probably do that. Basically, no matter what the world ended up being, I would still be fascinated by systems (whether biological or electronic or virtual) and would spend my time examining, learning and discovering about them.
Q. Can you talk about the local tech scene and developer community?
Though tiny, Singapore has a vibrant tech scene filled with startups and the big giants.
The government and the universities are highly supportive of startups and provide university scholarships, friendly immigration policies, seed funding and support of all kinds. The type of startups range the entire gamut from fashion e-commerce (Zalora to AI like nuTonomy with their self-driving cars.
Google is also here, building their Tango augmented reality system and their mobile experiences. Apart from the big names in software, there are a lot of fin tech companies since Singapore is a financial center for Asia.
A lot of the tech conferences in Asia happen in Singapore and I try to attend as many of them as I can.
There are meetup groups for all the technologies and I regularly participate in GPU meetups that happen here.
Q. What does your dream job look like?
My current work is pretty close to what I can imagine my perfect job to be. First and foremost, is working with great people.
People who are extremely intelligent and geeky, some of the smartest minds in the world, but who are also genuinely fun to be with.
Work is where we spend most of our adult life, so being with people I can respect and learn from is highly important to me.
Solving some of the most interesting problems in AI right now is amazing. Having the freedom to pick the languages, libraries and systems necessary for solving problems is important too.
Q. How do you normally find work?
The few jobs I have worked at were all through friends or plain dumb luck.
Usually, a friend who I highly respect and admire is working at a great place and I come to know of openings there. That is what has worked for me.
I have used job listings on LinkedIn and such, but they never quite worked out in the end.
Q. Leave us with a rant!
This is easy. The tech press and mainstream media have to stop their obsession and misinformation about AI!
Advances in GPU parallelism and cloud computing have scaled some classic neural network algorithms and methods to a point where they can be used to solve real world problems now. But that is all it is.
I have to roll my eyes every day reading bombastic claims and prophecies that are shared around. I would not be surprised if there is another AI boom and bust fuelled by all this hype.
If you’ve scrolled this far…
…thank you! We hope you enjoyed reading about Ashwin’s work and the Singapore / AI tech landscape — in that case, you should follow Ashwin on Twitter, check out his GitHub profile or read his blog — Code Yarns