Imon Bashir has been part of the MMT data engineering team as a data integration manager since 2018. In our interview, he gives us some insight into his daily work.
Name: Imon Bashir |
When I was a student, my studies comprised four areas: Mathematics, Software Development, Business Administration, and Information Systems. While it is hard to find a job that is solely built upon Mathematics, all students had the choice to go for any of the other three paths in their electives. The reason why I decided to go for the “Data” route (Information Systems) was that it felt to be the most diversified of the three. The process of working with data consists of multiple steps that include basically all aspects of my studies partially: Talking to clients and engineering requirements, building that in iterative cycles while programming with both database languages and programming languages.
Using all the skills I was taught and like without having repetitive tasks was what I desired and what jobs in this area - that have a wide variety of names - offer. As I have just foreshadowed, the names of this kind of work differ from company to company. So I applied at MMT where this was called “Data Integration Management” because from the description I felt like this was the kind of job that consists of the things I mentioned above. I have also had interviews with other companies, but from this interview more than any other, I became convinced that I am getting both the kind of work I wish and the kind of flexible, employee-friendly company I would like to work at and luckily, I got this employment.
My focus was on finding a company that allowed me to work with databases in a way that would also let me use my programming skills which I found. Another factor was the fact that having tremendous flexibility in terms of our preferred working style is more than just a buzzword MMT – it’s a reality we live in every day. At MMT, I was not only able to combine my studies with my job easily, but also able to combine it with private needs whenever necessary.
As a data integration manager, I’m responsible for dataflows between external APIs on the one side and the people that need that data on the other side. Defining what a client needs from a technical perspective and obtaining, transforming, importing, delivering, and maintaining that data – my role incorporates all of these activities.
This essentially means: I am initially talking to a client in order to understand their needs requiring both Business and Software Engineering skills on my end. Then, in the Data Engineering part, a mix of programming and database knowledge is needed: First, Programming is important in order to transform the data into the desired shape. Second, database knowledge is crucial for delivering data to the client in a way that is efficient on a long-term basis. It might be cheaper to import the data in some way, but an ill-considered import process only produces unnecessary additional effort in the long run. Sometimes, this is followed by a subsequent data analysis that combines programming and mathematics (mostly statistics).
First and foremost, it’s being able to be expressive in SQL and one additional programming language (Python being more helpful than R, but not necessary). Most problems break down into a transforming task and therefore many problems cannot be solved without being able to write queries that go beyond “SELECT * FROM”.
Being structured is essential. There are many different tasks that can pop up on the same day. Prioritizing and structuring them is crucial when it comes to solving these issues in a reasonable amount of time. Experience is another major factor because most of the time we only see that the integration is not working at the client-end. Knowing where problems usually lie for each data source or even knowing the exact problem can save a lot of time. I enjoy tackling these tasks because sometimes they are like little puzzles that need to be solved.
I could talk a lot about why data is important for companies, but that would be too general. A lot of dashboards, business decisions, and processes rely fully on the data we provide. If our team was not available and something were to crash in that time window, the client would have a huge problem.
To be honest, it’s never been about “that one project” but more about the diversity of what we do. From connecting the use of R or Python to using dozens of different APIs and UIs, there is always something new when it comes to working with an API for the first time. Or for example when already integrated APIs randomly start to change things without sending out notifications beforehand...
I think I would switch back to working as a data analyst, a job I did prior to this one. It’s pretty similar but different in that a data analyst uses as input the data that I as a data integration manager create as my output. That means I would stop providing data and instead use data and create insights from that data for stakeholders.