The true impact of AI based technologies can be felt in the healthcare industry where assistive robots are helping humans (particularly the elderly or disabled) with their daily tasks !
Krati Saxena, our next pathbreaker, AI Researcher at TCS, works on solutions for real-world problems in the textual domain using natural language processing, machine learning, and deep learning.
Krati talks to Shyam Krishnamurthy from The Interview Portal about her work with assistive robots in Japan, and the amazing things AI could achieve in numerous sectors that inspired her to pursue a master’s degree in Assistive Robotics.
For students, always strive to be a part of a team of people who come together to address challenges using cutting-edge technology for social good.
Krati, Your background?
I was born and raised in a modest Madhya Pradesh town. We were a middle-class family with ordinary earnings, though my parents made sure that I had access to every resource that might help me create my future. I was enrolled in an English medium school from the beginning. I was an extremely brilliant student who always received high grades in my exams. I was never athletic, but I was constantly involved in cultural and educational events like olympiads, quizzes, essay writing competitions, science fairs, and so on. I used to like participating in scientific displays the most, out of all of the activities. My project was chosen at the national level while I was in ninth grade, and it was the first event that piqued my interest in science and a little towards research.
After learning more about science-related occupations, I opted to take the IIT-JEE exam, which grants admission to India’s most prominent technical universities, the IITs. For my IIT coaching, my parents chose to relocate to a larger city. Unfortunately, I was unable to pass the exam on my first attempt but was able to do it on my second attempt after putting in a lot of effort, with discipline and commitment. Passing the exam allowed me to pursue my chosen professional path.
What did you do for graduation/post graduation?
I graduated from IIT Jodhpur with a BTech in System Science. I obtained not just technical knowledge but also life skills, thanks to state-of-the-art labs and equipment, excellent lecturers, and very bright classmates. I took courses in computer systems, electrical systems, mechanical systems, financial management, robotics, artificial intelligence (AI), and other subjects at college. I had the chance to perform research internships at IIT Kharagpur and NAIST in Japan, which sparked my curiosity in research, particularly in AI technology.
The project I worked on in NAIST Japan was in the healthcare domain, and it was a lot of fun. Countries with an aging population, such as Japan, have a smaller working population. As a result, most mundane jobs are mechanized by machines and robots. They also utilize a variety of technologies and systems to monitor older persons with ailments in their homes. Doctors can gain insight into suspected bone and joint disorders by tracking the patient’s body mobility (even if the patient is in a remote place or at home). Typically, motion capture devices are costly and difficult to set up in a realistic setting. As a result, we presented a method that uses a low-cost gaming gadget, Microsoft Kinect, to track the patient’s mobility. We employed artificial intelligence to train the system using real-world motion-capture data and then map the results on the Kinect device. We also used to go to guest lectures and seminars given by other lab members and academicians to learn more about their AI research. Seeing the amazing things AI can achieve in numerous sectors convinced me to pursue a career in AI. This internship was a significant incident that motivated me to improve my AI and machine learning skills.
Even though competing with some of India’s brightest minds (my batchmates) was difficult, I became the branch topper and silver medalist at the end of my graduation. I could do this with consistent attendance in lectures, disciplined study, and by maintaining the perfect balance between my studies and extracurricular activities. On the extracurricular front, I participated in a variety of sports, cultural groups, fest management teams, counseling teams, and other activities. All of these activities boosted my self-esteem and helped me develop social skills.
Later, I did my Master’s in Global Advanced Assistive Robotics from Kyushu Institute of Technology, Japan. These two years gave me a significant push towards research in AI. Living in Japan was a great experience and helped me learn a new language, new culture, time-management skills, and broaden my mind, skills, and network. After completing my education, I joined TCS Research. I am currently working as a Research Scientist here.
What made you choose such an offbeat, unconventional and uncommon career?
Science has always piqued my curiosity. As a result, I was sure that I would pursue a science career later in life. My instructors and friends advised me on all of the science-related job options available to me, as well as how to get here.
How did you plan the steps to get into the career you wanted? Or how did you make a transition to a new career? Tell us about your career path
A plethora of options arose as a result of my admission to IIT Jodhpur. In my second year, I applied for a variety of internships, including research-based internships, coding-based internships, management-related internships, and teaching-related internships.
At the time, I had two internship opportunities: a research internship at IIT Kharagpur and a teaching and management internship at TeachIndia. I chose IIT KGP because I believed that the research project will lead to more relevant opportunities in the future. This was a good choice because I received an international internship in Japan later in my third year. The initiative was about artificial intelligence in healthcare. I appreciated working on the project and being in the lab, where brilliant people came together to address challenges using cutting-edge technology for social good. I appeared in the placements in my last semester of college and received offers from Cognizant and TCS. TCS was my choice, despite the fact that my salary was lesser than that at Cognizant because the position was research-oriented.
I joined TCS in 2015. I had plans for higher studies in the back of my mind while I was pursuing BTech., so I decided to apply for colleges (outside India) in 2015. I gave GRE and TOEFL examinations. I got calls from a few colleges in Europe including Imperial College and the University of Edinburgh. With my contacts through my internships, I also got the prestigious Japanese government MEXT scholarship and a call from KyuTech (Kyushu Institute of Technology) in Japan. I decided to let go of European universities because those were 1- year master’s programs without any research projects. Moreover, they had no scholarships and charged hefty fees. Finally, I joined KyuTech as a Master’s student in Advanced Assistive Robotics.
I’ll share some personal observations on Japan. Japan has an aging population. That indicates there are more senior individuals in the country than younger people. As a result, there is a decrease in the working population. So many jobs in Japan are mechanized with the help of machines and robots. In India, for example, many individuals work as doormen, cleaners, security guards, waiters, and so on. Because these tasks are mechanized in Japan, with automatic doors, cameras, and security systems, vacuum cleaning robots, restaurant robots that take your orders, and other technologies, you will not find or seldom encounter individuals doing these jobs in Japan. Robots are also widely employed in the healthcare industry. Our lab was working on assistive robots, which is a branch of robotics that aims to help humans (particularly the elderly or disabled) with daily tasks such as eating, moving around, dressing, and using the restroom. These actions appear to be simple for humans, but are extremely difficult for robots. I was a part of a project that involved dressing a patient.
Robotics is a field that heavily relies on AI. If you’re interested in robots or artificial intelligence, Japan is one of the greatest places to visit. Although most Indians prefer to study in the United States or the United Kingdom, I believe Japan is also an excellent alternative for AI-related studies, because you can use AI and robots in your everyday life there, as well as work on some difficult real-world challenges that may benefit society in the long run. Obtaining the necessary materials for your research is also never an issue. In our lab, we used to have a variety of sensors, robots, equipment, and computing systems. And, unlike sharing everything with a group of individuals in India, we could use them individually as required. Even in older IITs in India, I have never seen such resources. The classes are taught in English for international students, and there are students from many different nations. The curriculum appears to be relatively simple for Indians, but the major focus is on research, which is often lacking in Indian colleges. Students learn a lot of research skills and practical experience as a result of this, which is very useful in the job market. .
After my masters, I joined TCS back in their research wing. I am still continuing with AI-related research in my job.
How did you get your first break?
I explored and applied myself for the first internship in my second year. Other internships and Job offers came through college internship and placement cells.
What were some of the challenges you faced? How did you address them?
Challenge 1: Securing a decent job and internship needed outstanding academic achievement and a thorough understanding of the field I was interested in (artificial intelligence). I had to keep up with theoretical knowledge and coding abilities because this was not just a software-related profession that required coding. This necessitated a lot of hard work and attention while balancing other college or career obligations. In my second year, I researched and applied for my first internship. Other internships and job offers came via internship and placement units at colleges.
Challenge 2: Keeping up with the technology is very important. Technology is something that is always evolving. Something that I learned while doing my BTech got updated when I went for Masters, and something that I learned in the Master’s course got updated in a few years, while I was on the job. So, learning never stops in the research field. With office work, I have to consistently take time out to read about new advances in technology.
Challenge 3: Here’s where I’d want to talk about mental health. When you don’t get a decent job or internship, or even when you don’t get high grades in an exam, it’s natural to feel unhappy. Other factors might also have an impact on one’s mental health. In the last five years, I’ve gone through similar periods several times. My family and friends, on the other hand, were always willing to help me with any issue I had. I learned to tackle obstacles with a grin after overcoming my concerns.
Where do you work now? What problems do you solve?
I work at the Pune office of TCS Research. We create solutions for real-world problems in the textual domains using natural language processing, machine learning, and deep learning. The job requires research acumen, knowledge of computer science, coding, and soft skills. I acquired these skills over the years while I was completing my degrees. I also keep myself updated with the latest research and technological breakthroughs by reading blogs, research papers, and tutorials, among other things. A typical day at work for me consists of meeting with the team to discuss new goals and the status of current projects. We spend the rest of the day working on the agreed objectives, which generally entail reading, producing reports and papers, and coding. I’ve always wanted to work in research. That’s why I enjoy working on projects that need me to bring new ideas and try different methodologies to achieve the best potential outcomes.
We work in a variety of fields. We’ve worked on Active-Learning based Annotation, Named Entity Recognition, and Information Retrieval, for example. In general, AI necessitates the development of training models, which demand lots of training data. The training data is made up of tags that have been annotated by humans. For example, we know that Google is a corporation name and California is a place in the line “Google is headquartered in California.” Humans tag such data in some format, which is referred to as training data, for a computer to learn this. However, in practice, we seldom receive labeled data, and manually labeling it is often impractical. As a result, we employ NLP and machine learning to swiftly annotate data and train models. After this, the models are utilized to get information from new data. We also concentrate on developing domain-specific dataset recommendation algorithms. A legal recommendation system has already been established. It collects the user’s information on their case and suggests activities to follow to get the best result. We also develop new formulated items, such as cosmetics. Experts at cosmetics firms must study a lot of books, papers, and internet information to learn how to make new cosmetics. They also need a significant amount of time and effort to create new versions of the formula by researching and evaluating existing data. We suggested a technique for generating variants that encompasses all data and analysis. In addition, it assists professionals by saving time and money while developing new formulae.
How does your work benefit society?
The projects we work on help people solve the problems with minimum effort. Our solutions help save time and money for various business domains due to automated solutions.
Tell us an example of a specific memorable work you did that is very close to you!
My Master’s work has left an indelible impression on me. That was the first time I worked directly with a robot, and I learned a lot about software, hardware, and solid research techniques as a result. The project involved detecting grabbing points on garments for the robot to conduct clothing assistance duties. My lab colleagues were working on a project that included using a robot to dress a patient, but the fabric had to be physically attached to the robot’s hand. To automate this process, I created image data and a deep learning system to predict which point of cloth to grasp and how to manipulate it to get the cloth in the required configuration. There, I was able to write my first conference papers. I attended conferences and local exhibitions which was an exhilarating experience for me.
Your advice to students based on your experience?
1. Do your own research: explore what you like, set goals and find multiple ways to accomplish your goals. If you are a school student, explore all the exams you can give, find all the relevant colleges and courses that may help you achieve your goals. If you are in college, start exploring job boards in your initial years to be aware of the requirements of your dream jobs, so that you can choose relevant courses while studying, as well as do online courses and certifications and enhance your skill set. Parallelly, also explore options for higher studies: what are the exams for various courses, which college to choose, which courses to choose in particular colleges.
2. Study: I cannot emphasize this enough: studying and practicing are the only ways to improve your skills and ensure academic achievement. Maintain a balance between your education and other obligations by engaging in extracurricular activities. And, in addition to studying, take part in mock tests and mock interviews since the only way to know where you stand and where you need to improve is to put your talents to the test.
I intend to continue working for a while. But, at some point in the future, I’d like to seek a PhD. Until then, I’ll continue to learn new technology and execute my work. Thanks!