BI & Data Warehouse Analyst

We're Hiring: BI & Data Warehouse Analyst | RAIA GRID

RAIA GRID, a subsidiary of Indosat Ooredoo Hutchison, is looking for a BI & Data Warehouse Analyst to build and manage our analytical data foundation and turn data into reliable business insights.

In this role, you will design and maintain our data warehouse, develop data pipelines, and deliver reporting that teams can trust and use. You will also apply AI and machine learning to support forecasting, anomaly detection, and natural-language querying, while building the data foundation these capabilities need.

Let's be part of smarter network!

Job Description

Requirement definition and metric ownership

• Work with business owners to turn vague reporting requests into defined metrics, each with an agreed calculation, an owner, and a stated purpose.
• Maintain a single business glossary so terms such as active customer, churn, utilisation, or revenue mean one thing across every report.
• Challenge requests that duplicate existing reports, and retire dashboards nobody opens.

Data warehouse design and development

• Design and maintain dimensional models — facts, dimensions, and slowly changing dimensions — that survive changes in the source systems behind them.
• Maintain a layered warehouse structure with clear rules about what belongs in staging, integration, and presentation, and enforce those rules.
• Model data from BSS, OSS, network inventory, and finance systems into one coherent view, rather than mirroring source schemas into the warehouse.

Pipeline development and operation

• Build and schedule ELT pipelines from operational sources, including incremental loads and handling of late-arriving or corrected data.
• Monitor pipeline runs, investigate failures, and reprocess without creating duplicates or silently dropping records.
• Keep transformation logic in version control and under test, not scattered across ad hoc scripts.

Reporting and self-service enablement

• Build reports and dashboards that answer a specific decision, each with a named audience and a defined refresh frequency.
• Tune semantic models and queries so reports open quickly enough that people actually use them.
• Prepare governed datasets so business teams can answer their own questions, instead of building every report on request.

AI-enabled analytics

• Identify where AI genuinely improves an analytics outcome — forecasting, anomaly detection, classification of free-text, natural-language querying — and where conventional reporting already answers the question.
• Deliver AI-assisted analytics use cases end to end, from problem definition through to a result the business acts on.
• Build and maintain the data foundation AI depends on: clean, documented, well-modelled datasets and a semantic layer that lets a model query them reliably.
• Establish how output quality is measured before anyone is asked to trust it, including a baseline to compare against.
• Work with the AI automation function on shared standards and tooling, so the company does not end up with two conflicting approaches to AI.

Data quality and governance

• Define and automate data quality checks covering completeness, uniqueness, referential integrity, and reconciliation against source systems.
• Publish lineage so that any number in a report can be traced back to where it came from.
• Apply access controls and the requirements of UU PDP 27/2022 wherever personal data is involved, including masking in non-production environments.

Stakeholder engagement

• Present findings honestly to non-technical audiences, including what the data cannot tell them and how confident the numbers are.
• Investigate and resolve disputes about the numbers — where two reports disagree, you are the one who finds out why.
• Document models and metric definitions so the knowledge does not live only in one person's head.

Job Requirements

We care that you can build a warehouse people trust, not that you have used a specific product. Equivalent tools and experience are welcome.

• SQL: Advanced — window functions, CTEs, query tuning, and the ability to read an execution plan rather than guess at performance.
• Data modelling: Kimball dimensional modelling to a production standard; familiarity with Data Vault or wide-table approaches and a view on when each fits.
• ELT and transformation: At least one framework such as dbt, Airflow, Azure Data Factory, SSIS, Talend, Pentaho, or equivalent.
• BI platforms: Power BI, Tableau, Looker, Metabase or equivalent — including semantic modelling such as DAX or LookML, not only building charts. Power BI is directly relevant to our environment.
• Databases: Relational databases, plus a columnar or MPP warehouse such as BigQuery, Snowflake, Redshift, Synapse, or ClickHouse.
• Programming: Python for data work; Git for everything that is not a one-off query.
• Orchestration and operations: Scheduling, dependency management, monitoring, and alerting on pipeline failures.
• Cloud: At least one major cloud, with an understanding of how storage and compute costs behave as data volume grows.
• Data quality: A concrete method for testing and monitoring data correctness. This is a requirement, not a nice-to-have.
• Machine learning: Applied ML for analytics problems — forecasting, classification, anomaly detection — using standard libraries such as scikit-learn, statsmodels, or Prophet.
• LLMs in analytics: Hands-on with at least one LLM provider API, including experience exposing warehouse data to a model safely via semantic layer, generated SQL, or retrieval over documentation.
• AI system implementation: Experience taking an AI component from prototype to something that runs on a schedule, is monitored, and has an owner — covering deployment, versioning, and refresh or retraining strategy.

Valuable, not required: streaming pipelines with Kafka; data catalogue and lineage tooling; geospatial and GIS concepts in a network context; telecommunications domain data; finance reporting experience; MLOps practices; vector stores and embedding-based search; semantic layer tooling such as dbt Semantic Layer or Cube.

Experience & Qualifications

• 5–8 years in business intelligence, data warehousing, or analytics engineering. Equivalent practical experience is accepted in place of a specific degree.
• Minimum 3 years building and maintaining a data warehouse in production — not only reporting on top of one that someone else built.
• Hands-on dimensional modelling experience, with a considered view on when to depart from it.
• Ownership of at least one reporting platform used regularly by business users, including handling complaints when the numbers looked wrong.
• Experience reconciling data across several operational source systems that disagreed with each other.
• Delivered at least one AI or machine learning use case that reached regular business use — for example capacity forecasting, anomaly detection, text classification, or natural-language querying over warehouse data — and able to explain how accuracy was measured and what it was compared against.
• Practical experience using large language models in an analytics workflow, with a clear view of where they add nothing and a simpler method is better.
• Experience in telecommunications, utilities, or another asset-heavy industry is an advantage, not a requirement.