Petrosea

Agus Jamaludin

A data scientist driven by insatiable intellectual curiosity, adept at uncovering valuable insights hidden within vast and complex datasets—whether structured, semi-structured, or unstructured. Skilled in applying mathematics, statistics, and data visualization to transform raw data into meaningful, actionable knowledge.

In an exclusive interview with CIOReview APAC, Agus Jamaludin shared his views on leveraging mathematics, AI, and data science for impact.

Career Growth Through the Lens of Mathematics

My background in mathematics has deeply influenced the way I approach problems, viewing them as patterns to be understood, modeled, and logically solved. The mathematical skills I developed during my academic journey have played a critical role in shaping my career as a data science professional.

Statistics, for example, provides the foundation for nearly all data science use cases. Once a problem statement is defined and business requirements are gathered, we collect and prepare the data, conduct exploratory analysis, apply descriptive and diagnostic analytics, and develop advanced models. These may include traditional statistical methods such as linear regression, time series analysis, and forecasting, or modern approaches using machine learning and AI. All steps lead to actionable insights that address the original goals.

Beyond statistics, my training included algorithm design, data structures, and analytical problem solving, shaping how I write efficient, scalable, and maintainable code. Designing and analyzing algorithms taught me to be strategic and systematic in building data-driven solutions.

Mathematical thinking, with its emphasis on precision, structure, logic, and data orientation, has also influenced how I communicate technical concepts to non-technical stakeholders. This ability to bridge business and technical teams has been vital in leading data initiatives and mentoring junior team members. Mathematics continues to serve as the backbone of my analytical mindset and approach to problem solving.

Challenges and Solutions

During our data warehouse migration projects, one of the main challenges I faced was ensuring data consistency and integrity while transitioning from legacy systems to a modern cloud-based platform. These projects were complex because they involved not only moving data but also preserving the logic behind data transformations, reports, and workflows. It required close coordination between business users, developers, and IT infrastructure teams.

We migrated our enterprise data warehouse, which integrates data from various sources and feeds dashboards and reports in Power BI consumed by departments across the business.

To ensure data integrity, we implemented scheduled monitoring and validation checks for both initial full data loads and incremental updates. We verified that each stage of the data process was running as expected and delivering valid outputs.

"Mathematical thinking, with its emphasis on precision, structure, logic, and data orientation, has also influenced how I communicate technical concepts to non-technical stakeholders."

Data validation was particularly challenging due to legacy logic built by third-party vendors and the departure of several original data owners and developers. To manage this, we initially limited the migration scope to the most critical elements, ensuring stability first, and then refined and improved the remaining elements after go-live.

One of the most significant outcomes of migrating from on-premises to cloud-based systems was the dramatic improvement in performance. Data processing became faster, system stability improved, and report generation times were reduced from hours to minutes, greatly enhancing operational efficiency and decision-making speed.

Predictive Analytics That Transform Business Outcomes

In my career, I’ve developed several predictive models and machine learning applications that delivered measurable business impact. One of the most recent projects was a GenAI-powered solution for Corrective Action Requests (CAR), which supports my company’s strong commitment to Safety, Health, and Environment (SHE) with the goal of zero harm. CAR reports potential hazards, incidents, and safe or unsafe behaviors in the workplace, and our model was designed to assist workers by increasing efficiency, reducing errors, and enabling faster, more accurate hazard identification and response. The AI automatically generates immediate control suggestions for hazard observations, categorizes reports by location, observation type, and activity, and even recommends long-term corrective actions such as risk elimination or substitution, appropriate protective equipment, and necessary administrative measures. By implementing this solution, we achieved a 70% improvement in reporting speed, a 40% increase in reporting volume due to the easier process, and a notable improvement in data accuracy, which the SHE team now leverages for strategic decision-making and stronger safety programs. This project not only enhanced operational safety but also reinforced a proactive reporting culture across the organization.

Generative AI and the Future of Industrial Data Science

Generative AI and other emerging technologies are significantly reshaping the role and capabilities of data science professionals across industries such as mining, energy, and infrastructure. These tools are not just automating tasks but are becoming strategic partners in problem-solving, enabling engineers and business users to be more effective and innovative.

One of the most transformative aspects of Generative AI is its ability to assist users across the entire project lifecycle, from ideation and use case discovery to development and implementation. For example, Generative AI can help engineers and analysts identify relevant use cases by analyzing trends, historical data, and business pain points. It can then translate these ideas into prototypes by generating code, queries, or workflows from natural language prompts, significantly speeding up development and lowering the barrier for non-technical users.

In my current company, we have seen tangible benefits from integrating Generative AI:

Document Understanding and Contract Intelligence

Generative AI helps teams summarize, extract, and interpret contracts and other legal documents, allowing departments such as legal, procurement, and operations to quickly understand key clauses, obligations, and risks. This improves efficiency and compliance.

Safety, Health, and Environment (SHE) – Safecard Automation

Generative AI enhances safety monitoring by classifying hazard observations and behavior-based safety reports, identifying causes of unsafe conditions, and recommending corrective actions. This increases accuracy, consistency, and proactive risk mitigation.

Enhancing Developer Productivity

For engineers, AI automates repetitive coding tasks, including code completion, SQL generation, and documentation writing. Teams can focus on solving business problems while the AI generates complex scripts or logic that would otherwise take hours.

What advice would you offer to individuals considering a career in data science today, especially given the rapid evolution of AI, machine learning, and emerging technologies?

For anyone considering a career in data science today, my main advice is to focus on building a strong foundation in problem-solving and critical thinking, not just in tools or programming languages. Technologies like AI, machine learning, and GenAI are evolving rapidly, but the ability to frame business problems, translate them into data questions, and deliver actionable insights will always remain valuable. At the same time, I encourage aspiring data scientists to develop a balance of technical and business skills—master core areas like statistics, machine learning, and data engineering, but also invest in domain knowledge and communication skills to create impact. It’s also important to embrace continuous learning; with AI advancing so quickly, staying updated through courses, open-source projects, and experimenting with new tools is essential. Finally, I would emphasize collaboration and adaptability: data science is rarely a solo effort, and the best outcomes come from working closely with cross-functional teams and adapting your approach as technologies and business needs evolve.