Petronas has built a unified enterprise platform, Enterprise Data Hub (EDH), to make data accessible across the organisation for managing and making use of its extensive data assets.
With over 300 subsidiaries and operations in over 100 countries, the company holds data across its integrated value chain, and the plan is to use this data as a business asset to drive strategic value.
It has selected Microsoft to aggregate data from multiple sources into a single unified platform.
Petronas’ head of enterprise data, Habsah Nordin, said EDH acts as a data factory that facilitates analytics across data domains and business entities.
It pools data from across the firm’s integrated value chain and has made data available, accessible, applied, and actionable to employees, she added.
According to the firm, it uses Azure Data Factory for data ingestion, and Azure Synapse Analytics and Azure Databricks to accelerate time to insights across data warehouse and big data systems.
EDH ingests data from 280 data sources, has 18 data domains, and integrates 506 data types.
Simplifying datasets search and discovery
Petronas launched a self-service portal and curated data marketplace, Data+, that allows employees to find and explore datasets from different domains.
It offers an intelligent search engine powered by natural language processing, automated tagging, personalisation, and a knowledge graph to enhance search accuracy and provide context for unstructured data.
An in-built chatbot to which users can request queries and get results further simplifies the process.
Nordin said an employee used to spend hours searching for information, which now happens immediately with just a few clicks, enabling employees to focus on getting work done.
Employees can visualise data by using Power BI in conjunction with Data+ to analyse and extract useful information from data and help in achieving better business outcomes.
The company is planning to introduce an unstructured data programme with Microsoft that will unify information across the enterprise using artificial intelligence and a large language model (LLM) to provide actionable insights on demand via embedded knowledge content analytics.
This will reduce the cognitive burden, mitigate risks, and empower employees to achieve operational excellence, Nordin said.
