Botnoi Voice
Sep 3, 2023
Label
UX/UI
What Are the Job Responsibilities of a Data Scientist?
So, you're curious about what a data scientist does? Let us break it down for you! Data scientists have a specific duty to add value to the business, and they do this in various ways: boosting revenue, cutting costs, spotting new opportunities, and minimizing waste. They use a bunch of cool tools and tech, both from inside and outside the company, to get the job done.
We often ask students what this career entails and usually get answers along the line to analyze data using AI. While this is correct, it is only half the story. A data scientist not only crunches numbers but also shares those insights with the right people to make smart decisions and drive the business forward. Think of them as data detectives who also need to be great storytellers! To get the whole picture, we need to also mention the following:
"A data scientist is responsible for leveraging advanced analytical tools and methods, including artificial intelligence, machine learning, and programming, to extract insights from complex data sets. This role involves not only analyzing data but also effectively communicating the results to relevant stakeholders and decision makers to ensure informed implementation and drive business value."
But other professions also add value to businesses, you see?
You're right. Other professions also add value to businesses. Using data as an example, a data analyst's role would indicate that the duties are the same, the goals are the same, but it's not necessary for it to be the same profession because the work processes or tools used may differ.
What Tools Do Data Scientists Use?
Data scientists have an arsenal of awesome tools that make them super effective. Besides being coding wizards, they need to know:
Statistics
Machine Learning
AI
Various visualization tools
And more recently, new technologies such as ChatGPT and LLM (Large Language Model) are becoming increasingly important as tools.
Data scientists must know how to use the appropriate tools that best align with their specific needs. Having a deep understanding of these powerful tools to determine which is most suitable for the job is key. Imagine you're a gamer playing a warrior character with a stash of epic weapons—you need to know when to use your sword or your magic spells to win the battle. You must know when and how to use the right tool in each situation.
Is It Always Necessary to Use AI and ML?
Not every problem needs AI or ML. Sometimes, basic stats can do the trick. Think of it like this: if you are a warrior fighting a small monster that’s at the bottom fodder, you don’t need to bring out your big guns. Just a simple punch will do. So, if the problem is straightforward, simple solutions can be more effective.
Give an Example of Segmentation Grouping
Let's say we have a problem that requires dividing student rooms. We can distribute them by distributing statistics, averages and standard values which are the same in every room. Companies do something similar to group customers into, say, gold or silver tiers based on spending, financial averages and lifespan.
But... say we want to divide customer groups. The same group of customers has different personality traits and similar consumption patterns. Data scientists can use the magic tool of machine learning to automatically cluster groups based on many factors.
Experiment with Segmentation Using Machine Learning
For the purpose of this example, we collected data from 40 students from an event that we did and then used K-Means clustering to divide them into 20 groups. Each group was divided into pairs and we ended up with pairs of besties that even a fortune teller couldn’t match better.
If we wanted bigger groups, we could tweak the settings to create cliques of four. For example, we ended up with four nice girls who all wore glasses. This method effectively turns us into a Sorting Hat that distributes students into different Hogwarts houses.
What Benefit Would This Have If Applied to an Organization?
Now, imagine this on a larger scale. For a huge company with thousands or even millions of customers, organizing them into neat groups can be a nightmare. But with AI, we can group similar people together, making it easier to tailor products and promotions to their needs. For marketing purposes, this is akin to having a superpower.
Summary of the Work of a Data Scientist
The real job of a data scientist involves understanding the problem, picking the right tools, analyzing the data, and most importantly, telling stories so that the leaders of the organization can understand them, which helps in decision making and adds value to the organization. Besides unsupervised learning like clustering, they also do supervised learning tasks like predicting customer behavior, spotting fraud, and forecasting trends.
Advice for Those Who Want to Be Good at Data Science
If you want to be a data science pro, we recommend you to get yourself used to answering questions using various tools and methods. You could start by finding problems and solving them on Kaggle, an online platform that collects a lot of data science problems.
We also recommend getting a co-op internship to get real-world experience. And hey, if you’re up for it, come join us at BOTNOI—we’d love to have you on the team!
กลับ
