从数据战略到生产流程:全面数据转型指南
Cloud data lake, data governance, industrialization for BI & data science
What we did
Challenge
- Structure and organize data and BI
- Power up data science capabilities
- Build IT best practices
Our approach
- Define data rules and process
- Deploy Cloud infrastructure
- Structure and train clients’ teams
Outcome
- Industrialized Azure solution
- Detailed Data governance
- BI and data science platform
Challenge
Our client is a global leader in B2B services related to client relationships. They approached us to help them structure and organize their data & business intelligence at the company level. Specifically they required a rationalisation of existing structures that would result in a unified company wide approach to data governance, data capture, roles and responsibilities, KPI definitions, usages, etc.While data is at the core of their activity, they had limited internal data science capabilities, with access to pre-aggregated datasets only. They wanted to improve these functions to better support their activity and growth in a more efficient way. They also wanted to accelerate development of Shadow-IT databases and usages.
Our approach
Data is produced and used by many teams in the company. To smooth the process of collection and analysis at scale, whilst maintaining clear value in a business context, we had to build synergies and create strong alignment between stakeholders. To do so, we designed and operated a collaborative development approach, in order to create a common set of Data Governance rules.To scale the project while continuing to improve readability and efficiency, we designed and deployed a Cloud infrastructure that would support user autonomy alongside sufficient levels of centralised control.Behind every data project and technological asset, there is a human. To ensure a smooth implementation, we also revamped the team structure, making sure we had the right people at the right place. We also spent time in building specific training programs aiming to increase data science literacy overall, and to guarantee a proper and documented handover.A crucial feature of our approach was the team’s engagement with a variety of stakeholders across the organization. From C-level executives to business managers, data scientists and IT staff across several countries, our team was able to gather insights and perspectives from a range of business and cultural backgrounds. This helped shape the language and communication of the project, as well as develop the business and technical aspects, forming a key factor in its overall success.
Outcome
For this project, we built a fully industrialized Azure solution, with every aspect defined-as-code and processed through Azure DevOps.We created a Data Governance Framework integrating the restraints of this specific business industry (global call centers etc) so that each data element has its adequate location in the IT landscape.We implemented a first set of use cases in an extensible framework, as well as integrating BI and data science capabilities into a single data platform.In successive engagements, our team designed, validated and implemented the data strategy successfully and to the client's full specification.