Team Updates

Behind The Scenes, Meet Our Data Scientist, Sahithya!

In this next edition of our Behind the Scenes series, we turn our attention to the fascinating technology of machine learning. Today, we speak to a member of our Machine Learning team, one of our data scientists Sahithya! She plays a crucial role in our machine learning operations, including driving the implementation of our recently released Date of Arrival predictions engine.

In this next edition of our Behind the Scenes series, we turn our attention to the fascinating technology of machine learning. Today, we speak to a member of our Machine Learning team, one of our data scientists Sahithya! She plays a crucial role in our machine learning operations, including driving the implementation of our recently released Date of Arrival predictions engine.

In this interview, Sahithya tells us more about herself, how she got intrigued with machine learning and some of her views on the profession within the logistics industry. 

Tell us more about you, who is Sahithya?

Sahithya: I’m a 25-year-old techie, who is happy to be in a profession which is also my passion. In my free time, you can find me practicing Indian classical songs in specific Carnatic vocal music. Quarantine has made me pick up some new hobbies like pencil-sketching (yup! I’m still bad at it :P)

How did you get started in machine learning? 

Sahithya: 3 years ago, I was working in a SaaS company where I saw a team build an automated chatbot for their product. The working of the bot which could mimic a human in their replies intrigued me. I became more and more interested after doing some research and seeing that mathematics was responsible for the construction of this bot. I’ve always loved mathematics and now understanding that mathematics and technology could solve problems was just cherry on the top. That’s how I started liking this field and how I became a Data Scientist. I also write about the field and contributed my first article recently for Towards Data Science, an international Data Science online publication.

What is your role in Parcel Perform? Tell us about your daily routine. 

Sahithya: Being a data scientist, my day-to-day responsibilities focus on understanding data within the Parcel Perform platform and working to comprehend the carrier or merchant’s parcel delivery and understanding the completeness of the data. 

A normal day starts with a company-wide regular stand-up, where we share the day’s tasks with the team. Then, I will start working on the Jupyter notebook to write codes to take raw data, understand it, and transform it into a format which the machine learning model can comprehend. After getting the results out of the model, it’s important to do an analysis of the results and whether it will add value to the business. I learn something new every day which is the best part. 

How does Machine Learning play a role in Parcel Perform’s operations? 

Sahithya: At Parcel Perform we’re passionate about the prospects of logistics data and what we can do. With refined logistics data, we can explore a lot of information and enhance the delivery experience of our customers. Machine learning is quite a new field for logistics, and we’re excited to explore possibilities of these technologies for not only our customers, but for the industry. 

What are some misconceptions about Machine Learning that you’ve encountered?

Sahithya: There’s a common misconception that machine learning systems can learn autonomously. This is not true. Machines need programmers to design learning algorithms for them to learn with tons of data. Only by feeding the model with lots of data and designing the learning algorithm for programmers to interpret the data for business use, the model will provide useful results. 

Parcel Perform has a strong partnership with Amazon Web Services (AWS’s) Machine Learning team. How has that collaboration influenced you as a Machine Learning professional? 

Sahithya: Machine learning is a broad field with several factors which needs to be considered on the modelling and data side to provide excellent results. Working with AWS is an exceptional experience, where I got to learn many data analysis techniques from the experts in the field. This helped to enhance my knowledge and perception on viewing the data which ultimately goes into the machine learning model that we’re building. Leveraging the experience of the AWS machine learning team and exchanging ideas with them exposed me to niche techniques and workable methodologies to perform error analysis and improve the performance of our Machine Learning model.

Our recently released product, the Date of Arrival Predictions engine is powered by Machine Learning. What was the primary motivation behind the creation of the Date of Arrival Predictions engine?  

Sahithya: When we order items from an e-commerce website, the range of dates given to us on when we might expect our parcel is often wide and not very meaningful. This is especially not a pleasant customer experience since with the use of machine learning to analyze with the underlying pattern and data, we can give exact information on the parcel journey. Another important motivation is, as a customer we aren’t informed of any issues that may arise during the delivery journey and it’s really important to provide regular updates. That was why we set out to build a prediction engine to address this gap in the delivery experience.

How would you describe the Date of Arrival Predictions engine in a single sentence? 

Sahithya: Elevate customer’s experience and retention by staying informed along the delivery journey. 

Find out more about the Date of Arrival predictions engine here

 

Want to check out who else is working with us? Read more about other team members at Parcel Perform on our blog, or go on to our LinkedIn page and get a glimpse of the people who make up our team.