Accenture’s Bilal Munir talks about how his role has changed with the growth of GenAI and highlights some of the most important skills he uses.
Bilal Munir has always been passionate about mathematics, which led him to a degree in actuarial mathematics at Dublin City University. While there, he landed an internship with Accenture in its data and AI practice.
“It was one of my career highlights – real project work, exposure to multiple industries and the ability to apply my coding and statistical skills from university,” he told SiliconRepublic.com.
After university, Munir rejoined the company through its graduate programme as a data science analyst in 2021. He then decided to specialise further in data and AI by completing a master’s degree in data analytics at University College Dublin. He now works as a data and AI manager for Accenture.
Tell us a bit about what brought you to your current role.
I initially joined Accenture’s global fraud and risk analytics team, so my first few roles were very counter-fraud focused. One of my first projects was around performing network analytics for a UK public sector body. I built and analysed the results of stakeholder interviews to form a directed network of the data-sharing initiatives between different government bodies to identify areas of improvement.
Later, I started working on local projects and spent two years working with a financial client. This role was more of a data engineering and cloud migration role as opposed to a data science role. I gained a lot of experience in areas such as data engineering and architecture, cloud computing and team management. These skills shaped me into a better data scientist and now I have an understanding of the end-to-end data lifecycle (from creation to ingestion and transformation through to deployment and monitoring) and it has given me a more holistic perspective when designing analytics and AI solutions.
After my two-year stint in the financial sector, I joined my current client, an Irish public sector body. I initially joined as the lead generative AI (GenAI) developer, bringing a small team of data scientists with me to help build and deploy GenAI prototypes for a small set of business users, proving their value while establishing our client’s AI strategy and responsible AI governance.
I spent the last few months of 2024 helping define our client’s responsible AI foundations, which is currently in an implementation phase. Nowadays, my main priority is to pave the path to production – to move the gen AI prototypes from testing into production for large-scale use.
How have you seen your role change with the explosion of GenAI in recent years?
Firstly, it has created a lot of demand – our clients want to know all about it and want to use it to solve problems that previously were either impossible to solve or they were too expensive to solve. I spend a lot of time with many of our clients in GenAI knowledge-sharing sessions – what it is and what it is not, what impact it will have on them and their industry, how they need to prepare for it and how to get started.
Secondly, it has been a wake-up call to those that want to deploy GenAI but can’t because they haven’t got the right data foundations in place – as we in Accenture call it, the ‘digital core’. Many companies that were previously slow or reluctant in cloud adoption have ramped up efforts in order to make use of the GenAI models that are served by cloud providers.
There is increased demand in existing work such as cloud migrations and data engineering, which then feeds into and enables the new demand in work relating to GenAI.
How do data professionals need to adapt in order to be ready for a GenAI world?
One skill that is common to most data professionals is coding. Code comprehension is one of the major benefits of large language models (LLMs). Developers can use GenAI in some shape or form to help with code issues and for code generation. These can be leveraged in multiple ways (a cloud provider, Microsoft Copilot, Github Copilot etc).
That being said, it’s still essential for the data professional to have a strong understanding of the code that is ultimately being used and to be able to accurately critique any code that is AI-generated – no LLM is perfect and overreliance on such tools at this stage will ultimately lead to issues. It’s not a magic solution – more of an additional tool in your toolkit.
Finally, alongside coding proficiency, all data professionals should strengthen their understanding of responsible AI and model governance, especially given the approaching deadlines of the EU AI Act.
What are the biggest challenges you face in your current role as AI continues to evolve?
One of the biggest challenges I’ve noticed is how much effort is required to keep up with the incredible speed at which both AI and GenAI is evolving. New tools, frameworks, models and methodologies seem to emerge every week, making it challenging to keep up.
It’s a great thing, but it also means that an approach you consider optimal today might be outdated in less than a couple months, requiring us to build with change in mind. I like to set aside a few hours each week to explore the latest tools and read the latest and most prominent research papers.
What do you enjoy most about your job?
The best parts of my job are the people I get to work with – our clients, our partners and our own people. Working with people who trust and support you – and whom you can trust as well – is absolutely essential.
Another major factor in my personal job satisfaction is flexibility – the ability to work in a hybrid setting really helps with managing my own work-life balance.
What are the most important skills to work in the area of data and AI?
Being the statistician that I am, I may be a bit biased when I say that a deep understanding of probability, statistics, linear algebra and calculus goes a long way. While you don’t necessarily need a deep dive into the mechanics behind every AI model to leverage them – just like you can drive a car without knowing how the engine works – having insight into how these models fundamentally operate helps you anticipate their strengths and limitations.
Beyond that, proficiency in coding and data manipulation is crucial. If you can’t access or prepare your data effectively, then success in AI will be difficult. Another key aspect is a continuous learning mindset, which helps with staying up to date with the latest industry trends.
What advice would you give to those considering a career in AI?
There are many areas one could focus on. I would suggest getting comfortable with Python and SQL, the go-to for most data professionals. Learn what the major cloud providers offer as you’ll more than likely be working with them in some shape or form.
And when it comes to the rapidly evolving realm of GenAI, get comfortable with the tools and frameworks that are being used: LangChain, LangGraph and Ragas are a few of the prominent ones.
Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.
