Harry Rafalimanana

Design and management of large-scale Power BI solutions (approx. 1.2 billion daily rows, approx. 600 users) for decision-making. Support for Product, Merchandising, and CRM teams, addressing performance, security, and international cloud migration challenges.

Key responsibilities and achievements

Deployment of end-to-end Power BI solutions for the Client Insights department, with full daily refresh on very large data models.

Preparation, transformation, and consolidation of data using Dataiku DSS (dataprep, processing, quality controls), upstream of analytical datasets.

Audit, redesign, and optimization of Power BI models and reports, with significant performance improvements (4 min → 20 s), removal of timeout limitations, and stabilization of data pipelines.

Implementation of the Power BI DATA chain:

-Functional and technical specifications

-Development and maintenance of reports and datasets

-Data and dashboard acceptance testing before production

-Refresh automation

Security and access management:

-Implementation of Row-Level Security (RLS)

-Access administration for approx. 600 monthly users across 13 Power BI reports

-Ensuring data compliance and confidentiality

-Analysis and resolution of incidents related to imports and refreshes (volume, performance, timeouts).

Migration of an On-Premise BI solution to Google Cloud Platform (GCP) and adaptation of solutions for China in the context of the Personal Information Protection Law (PIPL).

Design and implementation of a complete dashboard solution for the Lottery Business Unit during the migration of Business Object reporting to Power BI and Teradata.

Implementation of the complete DATA chain for Power BI reports:

- Framework and management of solutions

- Collection and formalization of needs from different Business Unit entities

- Drafting of technical and functional specifications

- Development and evolution of Power BI reports

- Acceptance testing of developments (data and dashboards) before publication

Construction of the datasets necessary for report implementation:

- Consolidation of multi-source data with high volumes

- Data reprocessing with a constant focus on performance optimization

- Calculation of indicators

- Implementation of automatic data feeding scripts

- Data quality audit

Business support:

- Support for business teams throughout the development process (from data approval to dashboard delivery)

- User training

- Performing ad-hoc analysis

Implementation of solutions for the Operational Risk department:

- Data Quality solution for monitoring and notifying on the quality of data shared between stakeholders and the operational risk community

- Decision support dashboard solution to facilitate steering and proactive risk assessment.

- Requirements gathering and solution prototyping

- Drafting of technical and functional specifications

- Analysis of the existing system

- Data audit and impact analysis

- Transcription of business rules

- Design of corrective patches

- Design and creation of Power BI Dashboards and KPIs

- Framework and management of solutions

- Design of data flows

- IT acceptance testing and business support