Career Nav #67: Applied Industry Research and Bringing Impact to Business
Shubhi Asthana, Senior Research Software Engineer at IBM Research and one of the Women Who Code nominees for 100 Technologists to Watch shares her talk, “ Applied Industry Research and Bringing Impact to Business.”
I've tried to be an impactful researcher as well as a developer. I've made some sustained contributions towards research and leadership in services computing, data science, and data engineering, which are important areas in the tech industry. What I like about applying this research is that it's not theoretical. You have a goal set in mind of what kind of client needs are there. You have to build up a competitive product, which is not readily available. There is a research angle to it. When we make an impact in terms of accuracy, performance, methods, and models, and the client is delighted, that is a success story.
In 2022, I started working on identifying objectionably, personally identifying information for IBM's premium machine learning platform, watsonx.ai. I worked on utilizing a novel classifier-based NLP approach that looks at the context of the potential spot in billions of unstructured documents and then trained it on LLMs with foundation models. While we did not build a completely novel algorithm, we utilized the different approaches of the algorithms, adjusting the decision bias to get the target recall and accuracy scores over vast data sets. There were more than 20 data sets and more than 20 data types.
We also developed a novel algorithm. We are tracking CO2 emissions. We're doing something towards the climate and working towards our LLM models. I've worked on varied projects like sales solutions, which received top honors at IBM. We were the finalists at the INFORMS Franz Edelman Award for Achievement in Advanced Analytics Operation Research and Management Sciences. The tool sets that we built for sales and solutions were top-class. They were highly valuable, and we saved a good amount of money by bringing targeted predictions on how our sales were going to be.
I've also worked on delivering innovative solutions that we've got more than, I think, 20-plus patents and more than 40 publications in top-tier conferences like NeurIPS, ITRO prolific data, ITRO police services, INFORMS, etcetera. For the innovative work that I have done, I got recognized internally and externally in the research and industry tracks. I received the Exemplary Technical Achievement IBM Outstanding Achievement Award five times. We also put in a project at conferences. We received first place in the ISSIP Excellence in Service Innovation Award. That was because of the business impact that our project had on the sales and solutions services.
I also received the 2021 Silicon Valley Business Journal's 40 under 40. My parents received the Suite Patent Recognition Award. I think they justify that the code we write or the innovative algorithms I come up with are not just impacting our clients but bringing in business value and revenue. Still, they're also ensuring that there is some give-back to society. There is a way folks are being motivated to work towards more impactful models, building more impactful projects. I like to work through professional organizations towards the advancement of careers for women.
Currently, I'm serving as the Industry Liaison Officer for Women in Operational Research and Management Sciences. I've been helping organize technical and career talks at the INFORMS business analytics conference and giving some Diversity and Inclusion talks. I'm also serving as the vice president of the local SWE section at Santa Clara Valley. I've managed to motivate women towards more leadership roles and invited some professional speakers who can guide women toward external eminence and recognition. I take my duties as the IEEE senior member very seriously by guiding them through the IEEE activities and IEEE Women in Engineering, and then also serving as a program committee in several conferences.
When I graduated from my undergrad and then my Master's, I didn't know there were so many options for engineering tracks to take. I started with robotics, went to wireless, software engineering, and then research-applied industrial engineering. The career path grows as we mold together and mold our personalities according to the teams where we fit in. I've always preferred the technical engineering track, contributing to the growth of the firm I am in, in this sector, with business performance and efficient delivery to the client's satisfaction. Those are the key goals that I have been pursuing.
I started a newspaper column called Tech Jargon in 2017. I wanted to introduce simple words like AI, "What is AI? What is machine learning? What is deep fake?" to those unfamiliar with the tech industry. We want to deconstruct and explain our tech sector of work in an easy-to-understand and grasp way. I get a lot of feedback on my columns from senior citizens who started with basic computers. They like the column because they now understand when somebody says artificial intelligence. It also encourages youngsters to take up STEM as a field of choice.
Apart from our routine research work, other challenges come in regarding changing our mindset with every new project we take on. Delivering out-of-the-box solutions can be tricky, especially when so many other folks are also building many innovative tools in this work stream. I've undertaken practical engineering applications in study and industry during my professional years. What interested me was the novel approach we can take to solve that issue: think out of the box.
It requires considerable literature review, trial and error, and intense brainstorming within the team to develop marketable results and enable us to patent the ideas we are getting. The key aspect I learned along the way is to be resilient and tenacious without losing focus on the desired result. Failure is part of the process. It teaches what is going wrong and where there is an opportunity to improve. Handling this helped me beyond the sphere of my professional environment. What motivates me further towards building more innovative solutions and working towards more accomplishments is I want to fortify and expand my projects to make an even bigger positive commercial impact in foundation models LLMs, especially in the space of NLP, the Natural Language Processing. I want to ensure that our clients are satisfied with our solutions. It involves co-developing with other business units within the firm, understanding the critical services requirements of the client, and then developing solutions to improve their productivity and revenue.