
Arnab Gupta, Data Analytics Lecturer at Business Technology Academy and a Lead Data & AI Consultant, recently spoke to us about his journey into data and how AI is changing the way people work with it. Arnab discussed how his career moved from programming and banking databases to reconciliation, data engineering, analytics and, eventually, consultancy. In the interview, he emphasised that understanding and checking the numbers is the key to understanding a business, and that while AI is a powerful tool, strong fundamentals and good judgement remain as important as ever. You can read the full interview below:
You started in programming. How did that lead you into the world of data consultancy?
Honestly, I didn’t start out with some grand plan to become a data consultant. I studied Electronics and Communication Engineering and always had a curiosity for technology. I’ve enjoyed learning things on my own, often just because I wanted to understand how they worked. I began in programming, then moved into a core banking solutions team where I worked with back-end databases such as DB2. I was writing SQL, investigating issues and trying to work out why certain numbers didn’t match. Somewhere along the way, I realised I could spend hours solving a data problem and actually enjoy every minute of it. That probably tells you something about me!
Then came a big personal change. I moved to Ireland to do my master’s in Data Analytics. While studying, I took a maternity-cover job at Planet, working in the accounts and reconciliation team. I was the only person there with an IT background; everyone around me was an accountant. They knew their processes inside out, but a lot of the work still involved paper, calculators and the same manual checks over and over. I remember thinking, surely we can make some of this easier. So I started building audit and reconciliation tools. I loved seeing my colleagues discover that a task they had always done manually could be done in a completely different way. What began as temporary cover eventually became a full-time reconciliation developer role.
Later, I joined Facebook’s Global Marketing Solutions (GMS) organisation, where I supported data scientists, programme managers and other teams working with advertisers. I worked on both the data engineering and analytics sides, and that opened my eyes to how much happens before anyone ever sees a report or a dashboard.
Eventually, I found my way into consultancy, and I can honestly say that’s where I’ve been happiest. I get to sit with people, understand their problem, work out where the data comes from, build the solution and see how it helps them. I don’t have to stay in one small corner of a project. Every project teaches me something new, and I really enjoy that.
When did you realise data was something you genuinely loved working with?
I’ve always loved numbers. But what surprised me was how calm I felt when I was working with them. Give me a messy database or a reconciliation that doesn’t add up, and my first instinct is to sit down and figure out why. Some people might find that stressful. I find it strangely relaxing! I think I realised quite early that I had a natural knack for spotting differences and understanding the story behind them.
There’s something I’ve carried with me through every role: if you can understand and reconcile the numbers, you can begin to understand the business. A figure on its own doesn’t mean much. You need to know where it came from, what it represents and why it changed. Of course, numbers can mislead when the underlying data is wrong, which is exactly why I care so much about checking them.
I get far more satisfaction from explaining a number properly than from making a dashboard look beautiful. When the numbers finally make sense, that’s a wonderful feeling. And when that understanding helps someone else do their job better, it’s even better.
What does a data consultant actually do, and where do SQL, Python and Power BI fit in?
Imagine someone comes to me and says, “Our reports don’t match, and nobody can tell us why.” I can’t solve that by opening Power BI and making a colourful chart. First, I need to listen, understand how their business works and follow the data back to its source. Sometimes the real problem is in the process, the definition of a metric or the way different systems communicate with each other. That’s the part of consultancy I love: understanding the whole problem and helping people work through it.
As for the tools, I see them as different parts of the same story. SQL lets me ask questions directly of the data and check what’s really there. Python is brilliant for taking repetitive work off our hands, cleaning data, automating processes and doing deeper analysis. Power BI helps me turn the results into something people can understand and use. I also work with cloud platforms such as Microsoft Fabric to bring the pieces together.
When I teach, I try to get students to see beyond the software. A tool is useful only if you know what you’re trying to solve. You can create the most impressive dashboard in the world, but if the number on it is wrong, what have you really achieved?
Everyone is talking about AI. How is it actually changing the way people work with data?
I’m genuinely excited about AI. It can help us explore an idea, draft a SQL query, find a problem in Python code or begin making sense of thousands of written survey responses. Work that used to take a lot of repetitive effort can often get a much quicker starting point. That leaves more time for understanding the problem and thinking creatively about the solution.
But here’s the thing I always want students to remember: AI can sound incredibly confident and still be completely wrong. If it gives you a number, don’t just fall in love with the answer. Ask where it came from! You still need to check the data, understand the business rules and protect confidential information. My background in reconciliation has made me quite stubborn about that, in a good way.
I don’t think the exciting part of AI is simply getting it to write code for us. It’s what we can do when we combine these tools with real understanding and good judgement. That’s why I still encourage students to learn SQL, Python and the fundamentals properly. You need to know enough to ask better questions and recognise when an answer doesn’t make sense.
What would you say to someone who is curious about data but isn’t sure they belong in tech?
First, I’d say you absolutely don’t have to fit one particular mould. I think back to my time at Planet, surrounded by accountants. They understood the business in ways I didn’t, and I could bring technical ideas they hadn’t worked with before. That combination was incredibly valuable. People coming from finance, retail, operations or completely different backgrounds often have more useful knowledge than they realise.
Start small and build confidence. Learn to ask questions with SQL, explore a dataset using Python and put together a Power BI report that tells a clear story. Choose a problem you’re genuinely interested in. And please don’t panic when the data is messy or the numbers don’t match. That’s often where the real learning begins!
My own career has hardly been a straight line. I’ve changed countries, studied while working, moved through banking, payments, data engineering, analytics and consultancy, and now I get to share what I’ve learned with students. One of the things I enjoy most about teaching is helping people see how many directions a career in data can take. It’s much bigger than the title “Data Analyst”. You might discover a passion for engineering, business intelligence, consultancy or AI.
You don’t need to know everything on day one. Even after more than ten years of working with data, I’m still learning new technologies and finding things I don’t know. That is part of what makes the work interesting. Stay curious, keep asking why, and don’t be afraid to try something new. The first time your work helps someone through a problem they’ve been struggling with, I think you’ll see why I love this field so much.
To learn more about Business Technology Academy’s Data Analytics Course you can check out the programme in the Courses page