AsianScientist (Mar. 27, 2019) – You can read about artificial intelligence (AI), blockchain and cloud computing in the news, or study them as the new ‘ABC’ in school, but there’s nothing quite like jumping right into the deep end of the tech industry and experiencing innovation at the frontline.
For Gabriel Wong, an undergraduate at the Singapore University of Technology and Design, participating in SGInnovate’s inaugural Summation Program was an ideal way to spend his summer break from May to August 2018.
Open to students and recent graduates of institutes of higher learning, this unique apprenticeship provides selected candidates an opportunity to ‘get their feet wet’ in deep tech projects. Over a period of three to six months, they will work alongside experienced software and engineering professionals at Singapore-based organisations to tackle real-world problems, such as distilling insights from Big Data.
“I major in computer science, with a minor in business analytics and AI. To me, programming is like solving puzzles, so I thought: why not try and solve a puzzle with real impact on industry or society?” he quipped.
Earning his stripes
Prior to applying for SGInnovate’s Summation Program, Gabriel had written to AI solutions firm TAIGER for an internship, drawn by its vision to “automate manual yet complex tasks, intelligently”. Back then, he was already dabbling in Python programming and had devised a prototype of a ‘smart mirror’ that would display news and weather information to a user, despite not having been exposed to AI in his coursework at school. Looking to increase his chances of securing the internship, he sent in another application to TAIGER through the Summation Program and was accepted.
Of the many flavours of AI that TAIGER was working on, Gabriel was assigned to a project that involved using computer vision and natural language processing to extract information from scanned documents.
“If you think about the amount of paperwork that organisations have to go through each day, a lot of time could be saved if a recognition system could automatically identify and digitally record information in forms,” he explained.
Granting machines even this level of intelligence was easier said than done. Gabriel spent a large part of his apprenticeship tweaking the parameters of a computer algorithm to make it more accurate in its interpretation of the things it ‘sees’. The learning curve was steep, given his relative inexperience with the technology.
“I had to pick up a lot of the concepts and practical skills from scratch,” he said. “Thankfully, my mentors and colleagues were very willing to share their knowledge and expertise with me.”
A trial by fire
But refining the AI-enabled recognition system in a sandbox environment was the easy part—the true stress test came when Gabriel had to deploy the AI in a real-world scenario.
“During development, I was able to achieve the primary objective of extracting specific types of information from paper documents. However, when field testing the system, there were circumstances that we did not foresee, such as time and language constraints,” he noted.
For example, scanning a document and allowing the AI software to read and record information took ten minutes—too long for client-facing operations that require a fast turnaround time.
“You can’t just say ‘wait’ and expect a client to stand by for ten minutes for your system to complete its task,” Gabriel mused. “These are the kinds of things you don’t really encounter in an academic environment but will need to consider when building commercial applications.”
To improve on his AI software, Gabriel had to repeatedly go back to the drawing board: like most agile development workflows in software engineering, there is no truly finished product, only one that undergoes continuous cycles of iteration and improvement, he explained.
The start of a journey into deep tech
Challenging as the process was, Gabriel found that exposure to these realities helped him better understand the demands and rigour of the field of deep tech. He enjoyed the experience at TAIGER so much that he continued to work part-time with them after he completed the Summation Program in August 2018.
“Through the Summation Program, I was able to get a very good idea of what it’s like to be a software or AI engineer. AI development is something that I will want to do again in the future,” he said.
For tech enthusiasts keen on participating in SGInnovate’s Summation Program, Gabriel emphasised that they should not be intimidated by a lack of experience.
“Of course, it helps if you have some knowledge of software engineering, or some personal tech-related projects that you’ve worked on to demonstrate your interest and passion. Having said that, when I applied to the program, I was also relatively new to AI, but I just decided to go for it, and it turned out for the best,” he said.
To continue to engage graduates of the Summation Program, SGInnovate will keep alumni such as Gabriel in the loop about deep tech events, courses, workshops and networking sessions. Subsequent cohorts can also look forward to foundational courses before commencing their apprenticeship, which will better prepare them for their roles at the deep tech organisations of their choice.
SGInnovate believes that human capital is a critical aspect of its Deep Tech Nexus strategy. Hence, it created the Summation Program, which is currently in its third run. The program will allow apprentices to work in technology-intensive companies on deep tech projects such as AI, machine learning, and deep learning.
Taiger is one of SGInnovate’s portfolio companies.
Asian Scientist Magazine is a content partner of the SGInnovate.
Copyright: SGInnovate. Read the original article here.
Disclaimer: This article does not necessarily reflect the views of AsianScientist or its staff.