Introduction
Data science has become increasingly important in today’s world due to the abundance of data generated daily. From social media to e-commerce, businesses generate large amounts of data that can be leveraged to gain insights and make informed decisions. Data science involves the use of statistical and machine learning techniques to analyze and make sense of this data. By uncovering patterns, trends, and insights, data scientists can help organizations optimize their operations, improve their products and services, and ultimately increase their bottom line. Data science’s importance will only grow as businesses and industries increasingly rely on data to drive their decision-making processes.
Today’s speaker Akshit Bhalla is one such skilled professional who has been able to leverage his expertise in the field of data science to solve a range of business problems. So, let’s learn about this revolutionary field from Akshit Bhalla.
Know More About Akshit Bhalla
Akshit Bhalla is a Data Scientist at HP Inc, Bangalore. He has gained nearly 3 years of experience in the field of Machine Learning. He comprehensively understands applying Machine Learning techniques to diverse domains such as Supply Chain Management and Customer Support. With expertise in Big Data and AWS/GCP based MLOps platforms, he has successfully led several projects and offered innovative solutions to various business problems. Akshit collaborates with data scientists and other professionals to identify problems and develop solutions to make informed decisions about a business/organization.
In this article you will get to know the following things:
- Understanding the importance of internships and their role in developing essential skills and gaining practical experience in data science and related fields.
- Gaining knowledge of the different domains and applications of machine learning, including supply chain management and customer support.
- Learning about the various types of data, including tabular datasets, images, and sequence data, and how to work with them using tools like AWS and GCP-based MLOps platforms.
- Understanding the role of software engineering skills in data science and how they can be leveraged to develop and deploy ML pipelines for real-world business problems.
- Developing effective communication skills to balance technical accuracy and rigor with clarity and accessibility when presenting data to non-technical audiences.
The Journey From an Intern to Data Scientist
AV: Good day, and welcome to Analytics Vidhya! Could you please introduce yourself and provide some insights into your educational and professional background?
Mr. Akshit: Thank you for having me. My name is Akshit Bhalla, and I am currently working as a Data Scientist at HP Inc. in Bangalore. I hold a bachelor’s degree in Industrial Engineering and Management from RV College of Engineering in Bangalore, which I obtained in 2020. With nearly three years of experience in the field, I have developed a comprehensive understanding of how Machine Learning can be applied to various domains, such as Supply Chain Management and Customer Support.
My expertise lies in working with diverse types of data, including large tabular datasets, images, and sequence data, such as text. Leveraging my proficiency in Big Data and AWS/GCP based MLOps platforms, I have successfully led several projects and brought innovative solutions to a plethora of business problems.
AV: That’s truly impressive! As someone with internship experience at Bosch, IIT Kharagpur, And Searce inc, Can you please elaborate on the skills you acquired during your time as an intern?
Mr. Akshit: Certainly. I interned at Bosch during the second year of my undergraduate studies and worked on Statistical Quality Control. It was my first time wrangling with real-world data that I collected from their manufacturing unit to apply my knowledge of statistics to trace manufacturing defects. Data collection, Statistics, and Quality Control were the top skills I learned.
The following year I ranked amongst the top 1% of students nationally in the examination on Data Mining by IIT Kharagpur and was offered a research internship at the same institute. I worked on Skyline Queries on Uncertain Graphs and learned the importance of patience and discipline required to conduct quality research. Also acquainted with the pace of ML research as well as how to publish papers.
Got an opportunity to intern at Searce Inc during the final semester of my undergraduate studies. I was introduced to Machine Learning on Google Cloud and AWS and learned to architect, develop, and deploy ML pipelines for real-world business problems. Also, worked with a variety of data like tabular, text, and images, as well as an array of technologies such as AutoML, APIs, and other cloud-based applications. Additionally, I was exposed to the service-based industry and its operations.
AV: That is truly remarkable! Can you please elaborate on how your internship experiences have influenced your personal and professional growth?
Mr. Akshit: My first internship introduced me to the corporate world, the second to research, and the third to technology. I knew very little about Data Science when I interned at Bosch. However, I developed a liking for statistics, which motivated me to pursue Data Science as a career. At IIT Kharagpur, I realized the rapid growth of the field and the necessity of discipline and persistence to keep up with it.
Consequently, I was well-equipped for the final year placements, and my projects’ quality, my accomplishments, and the “IIT tag” on my resume caught the attention of several companies offering Data Science roles. Ultimately, I decided to intern at Searce, where my professional journey began. I worked on a number of exciting projects there for the next 1.5 years, which laid a solid foundation in Data Science and Machine Learning. This facilitated my switch to a larger, product-based company, HP Inc.
AV: Your journey from a software engineer to a data scientist is fascinating. How your experience as a software engineer has helped you in your role as a data scientist? What specific skills that you learned as a software engineer have been instrumental in your current role?
Mr. Akshit: I launched my professional career at Searce Inc as a Software Engineer working on Cloud-based Machine Learning. I worked with the Applied AI team to build Google Cloud and AWS ML solutions. As someone without a background in Computer Science, my learning curve was steep, and I picked up quite a lot in a short span of time. After recognizing me as the “best performer” for several months, they promoted me to Data Scientist within the first year. Some essential skills that I picked up as a Software Engineer include – setting up event-based triggers, spinning up virtual machines, memory and storage management, as well as architecting and orchestrating big data pipelines on the cloud.
AV: Could you shed some light on your unique approach to data science and what sets you apart from other professionals in the field?
Mr. Akshit: Data Science was introduced to me through statistics, cultivating a profound appreciation for data-generating processes, distribution fitting, and hypothesis testing. This statistical knowledge has proven invaluable in my understanding of learning algorithms and techniques for diagnosing them when they fail to perform as expected. Likewise, my exposure to Computer Vision came through the perspective of Electrical Engineers and Linguists shaped my understanding of Natural Language Processing. Most individuals know about these fields purely as machine learning applications, hindering their ability to appreciate the elegance of data in the same way. This unique approach distinguishes me from others.
Furthermore, my background in Industrial Engineering and Management has provided me with a strong comprehension of business operations. This understanding has enabled me to tackle business challenges more effectively by applying Data Science to provide data-driven solutions.
AV: How do you balance the need for technical accuracy and rigor with the need for clarity and accessibility when presenting data to non-technical audiences?
Mr. Akshit: In my role as a Data Scientist, my primary focus is to solve business problems using data generated from business processes. For example, I utilize data at HP to optimize customer support operations. The ultimate goal is to deliver solutions that help the business, whether or not they involve machine learning.
While technical expertise is crucial, it is equally important to present my findings in a clear and concise manner, especially when addressing non-technical audiences such as top management and business leaders. To achieve this, I strike a balance between technical correctness and clarity by emphasizing the solution and its impact, rather than delving into technical details. I avoid using technical jargon as much as possible; when necessary, I provide simple definitions to accurately convey complex concepts.
To effectively communicate with non-technical audiences, I believe having a good understanding of both the business and the technology is essential. I enable clear and effective communication of my ideas, ensuring that those who are not technically inclined can easily understand and relate to my work. Ultimately, non-technical audiences, who prioritize value over technicality, consume my work.
AV: Knowing more about the person behind the professional title is always interesting. So, could you share with us some of your favourite hobbies or interests? How do they contribute to your personal and professional development?
Hobbies can help us develop skills like creativity, Problem-solving, And teamwork, among others. So, We’d love to hear about your hobbies and how they impact your life!
Mr. Akshit: In my free time, I am passionate about fitness and regularly go to the gym. I also enjoy gardening and playing basketball, and have been in sports from a young age. Through my participation in sports, I have developed various qualities, including a commitment to healthy living, initiative, planning and strategizing, leadership, and learning how to manage failure. These qualities have proved invaluable in my personal and professional growth, allowing me to approach challenges with resilience and determination.
For instance, when faced with job rejections or failed projects, I do not take them personally. Instead, I take ownership of my mistakes and analyze them to identify areas where I can improve. I firmly believe that it is essential to learn from failures and mistakes to grow both personally and professionally.
AV: How your experience working at HP has impacted your career development? What are your goals for your career in the next five years?
Mr. Akshit: Joining HP was the best career decision. My managers, Shashavali Godala and Bengu Altinordu, are amazing. They care about my professional development as much as my personal well-being. The company culture is very supportive, and my colleagues are always willing to help me when I have questions or need assistance. In terms of my career development, working at HP has given me the opportunity to develop my technical skills and business knowledge. I have had the chance to work on a variety of projects, which has allowed me to gain experience in different areas of the business. This has helped me to identify my strengths and weaknesses and to focus on areas where I need to improve.
Looking ahead, my goals for my career in the next five years include continuing to develop my skills and knowledge and taking on more challenging projects. I would also like to take on a leadership role within the company and mentor and train new employees. Additionally, I am willing in exploring opportunities to work in different parts of the world. I believe that exposure to different cultures and ways of working can help broaden my perspective and enhance my skills.
AV: Can you share with us a recent project you worked on and the key insights you gained from the experience?
This would give us a better understanding of how you approach and execute data science projects, And how you extract meaningful insights from data.
Mr. Akshit: In my role at HP Inc, my primary focus is working with sequence data, including text and telemetry signals. Currently, I lead the device failure prediction project that involves forecasting early life failures in LaserJet printers using telemetry signals. While I cannot share the actual implementation details of this project due to confidentiality, I can provide the top three takeaways that I have gained, which are applicable to almost any data science or machine learning project.
[A] Firstly, it is crucial to understand the business problem clearly before initiating development. Often, we may realize that we do not need machine learning to solve the problem or that we do not need to train a model because one is readily available online. It is also essential to figure out whether the required data will be available for the model at the prediction time. Having a clear understanding of the business requirements helps reduce the development effort significantly and results in better models built with domain knowledge.
[B] Secondly, prioritizing data quality is vital since the model is only as good as the data it sees. Real-world data can be messy and riddled with human errors, incompleteness, and inconsistencies. Cleaning, transforming, and merging disparate datasets for modeling requires patience, and it is essential to be careful about data leaks. It is true that more than 80% of the time spent on a project is used for preprocessing.
[C] Lastly, analytics, reporting, and dashboarding are valuable components of any data science project pipeline. While this type of work may be delegated to data analysts, it is equally crucial for data scientists to be well-versed in it. Conveying key insights to business leaders through Power BI or Tableau dashboards facilitates data-driven decision-making.
AV: With the field of data science evolving constantly, What technological advancement or trend fascinates you the most? What specifically about it interests you?
Mr Akshit: The emerging field of Meta-Learning excites me. Machine learning models can only learn tasks they were explicitly trained for. However, with the advancements in zero-shot and few-shot learning, the future of machines is exciting! The next generation of machines will be able to generalize to examples that they have not encountered before, which is real Artificial Intelligence in my opinion.
Just imagine a world where models can automatically generalize to novel domain-specific examples, without the need for any fine-tuning or training. Such technology will be a significant step towards the democratization of AI, as non-technical individuals will be able to utilize these models without any prior knowledge of technicalities.
AV: How do you see the field of data science evolving in the coming years? How do you plan to stay ahead of the curve in terms of skills and knowledge?
Mr Akshit: The pace of development in the field of ML/AI is incredibly fast, with new models, technologies, and products becoming popular each day. A few years ago, the term “NLP Engineer” was uncommon, but today, roles such as “Prompt Engineer” have become popular. With the advent of Multimodal Machine Learning and Large Language Models, it is clear that data scientists of the future will need to have a good understanding of both Natural Language Processing and Computer Vision.
To stay ahead of the curve, I make sure to have a solid grasp of the fundamentals of these two domains. Whenever a new model, such as YOLO-8 or ChatGPT, becomes popular, I take the time to read through its documentation and understand how it works, taking away key ideas. Additionally, I believe that companies like HuggingFace are doing an incredible job of making it easier for us to work with the latest models. This, in turn, frees up time to think about how emerging models will be useful in business, which keeps me ahead of the curve.
AV: Can you describe the initiatives or programs that HP has implemented to ensure that employees stay up-to-date with the latest technological advancements happening globally?
Mr. Akshit: HP Inc prioritizes the continuous education of its employees by providing frequent training and presentations on the latest technological advancements. Additionally, HP offers access to a vast collection of resources that are typically not available for free. Recently, HP collaborated with O’Reilly, which provided a wealth of books and high-quality video lectures for our use. HP also supports the pursuit of relevant certifications and professional development programs to enhance the skills of its employees.
AV: Looking back on your career as a data scientist, What are three things you are most proud of, And how have they helped you shape as a person?
Mr. Akshit: I take great pride in having started my journey with Statistics instead of jumping straight into Deep Learning like many do these days. Understanding the basics is crucial in the real world, where memory, latency, and storage space limitations can make a Deep Learning model impractical. The ability to choose a simple model is an underrated skill that I developed by having a strong foundation in Statistics and Machine Learning.
Secondly, I am proud of my firm and determined approach to decision-making. Despite pressure from my college’s placement cell, I was determined to become a Data Scientist and did not apply for any other jobs. Many people believed that as an Industrial Engineer, I could not become a Data Scientist, but I was happy to prove them wrong. My conviction has allowed me to accomplish a lot in life, and my move to HP was a result of my determination to be part of a greater community.
Lastly, networking has been critical to my success in Data Science. Back in college, I interviewed Sheela Siddappa, the Global Head of Data Science at Bosch at that time, and learned a great deal about Data Science in the real world. I continued to network on LinkedIn, seeking professional guidance and advice, and these efforts in the early days shaped my career trajectory. I cannot stress enough how important it is to build a strong network of professionals and seek advice from those with more experience.
AV: What advice would you give someone interested in pursuing a career in data science, And what skills or qualities do you believe are most important for success in this field?
Mr. Akshit: When individuals seeking advice on pursuing a career in Data Science approach me, I always suggest they start by familiarizing themselves with the various roles and corresponding responsibilities within the field, such as Data Analyst, Data Scientist, ML Engineer, Business Analyst, Data Engineer, and MLOps Engineer. While the hype surrounding Data Science and its relatively low barriers to entry may attract many, it’s important to enter the field with a genuine interest in solving problems with data rather than unrealistic salary expectations.
Too often, those who jump into Data Science without exploring all the potential job roles end up disappointed and either leave the field entirely or switch to a different data profession. I have personally seen friends who started as Data Scientists and ultimately chose to become Data Engineers because it better suited their interests and skills. Thus, exploring all job roles within the field is important before committing to Data Science.
I highly recommend starting with a strong foundation in basic statistics, such as linear models and probability distributions, for those who decide to pursue Data Science. As the field of data science and machine learning continues to rapidly evolve, it’s essential to have a solid understanding of the fundamentals and a constant thirst for learning in order to succeed.
AV: Thanks for your time! Definitely, This will be helpful for all the aspirants who want to go into data sciences or data fields. So, Let’s conclude today’s discussion.
Conclusion
In conclusion, Akshit Bhalla’s journey from an industrial engineering student to a data scientist at HP Inc inspires aspiring data scientists. With his ability to solve complex business problems, expertise in diverse data types and Big Data technologies, and his passion for learning and experimenting with new techniques, Akshit has shown that success in this field requires a combination of technical skills, patience, and discipline.
His experience also emphasizes the importance of internships, which helped him develop his skills and gain valuable industry exposure. Akshit’s journey is a testament to the limitless possibilities in data science and the impact it can have on businesses and society.
If you want to become a Data Scientist, click here to know more.
