Anne-Sophie Roessler, International Business Developer, Dataiku - "3 ways to Fail your Data Lab Implementation"
1 like501 views
The document discusses the implementation of a data lab using Dataiku DSS, highlighting its advantages such as streamlined workflows and the ability to address multiple use cases with diverse expertise. It emphasizes the importance of collaboration between technical and non-technical teams, and outlines a structured approach to data management, from acquisition to deployment. Key focus areas include governance, promoting diversity, and ensuring end-to-end project execution.
Anne-Sophie Roessler, International Business Developer, Dataiku - "3 ways to Fail your Data Lab Implementation"
- 1. Three ways to Fail your Data Lab Implementation
- 2. Dataiku DSS
- 3. DataLabs 10 M€ in 2014121 499 M€ in2014 3 029 M€ in2015 5 454 M€ in2014816 M€ in201410 M€ in 2008 Marketing/ Web ü Behavioral segmentation ü Churn prediction ü Sales forecast ü Dynamic Pricing Industrie& Infrastructure ü Predictive maintenance ü Logistic Optimization ü Smart Cities Bank & Insurance ü Fraud detection ü Riskanticipation ü Lifetime moment detection
- 4. Why a data Lab? • 1 single Workflow : from a segmentated workflow to a transversal one • Several use cases: Ability to adress many different data centric topics within a single unit • Multiple competences: Business focused approached mixing many different competences • End to end projects : combining data from different sources to handle several aspects on a single topic
- 5. Deployment ofthe predictions Dataiku DSSfor fraud prediction Client service Sensor data Garage data Administration • 1 Project Owner (IT) • 1 Project Manager (Business) • 1 Data scientist in house • 3 data scientist sfrom 3 different firms • 3 consultants from 3 different firms • 1 architect (external) Accepted file INVESTIGATE ! Thetransactions areblocked dependingontheir gap with the business rules and behavioral patterns
- 7. Focuson the framework,not on the input Data Acquisition & Understanding Data Preparation Model Creation Evaluation Deployment Scored dataset Scored dataset Iteration 1 Iteration 2 Iteration n ✓ Read and import raw data ✓ Detect schemas and structure ✓ Analyze distributions ✓ Assess quality: outliers, missing values... ✓ Performance metrics ✓ Robustness & generalization (cross validation) ✓ Insights (eg variable importance) ✓ Create derived and aggregated variables → Analytical dataset → Report ✓ Feature selection ✓ Compare algorithms ✓ Scoring engine ✓ Publish predictions ✓ Monitor performance ✓ API Business Understanding Adapted from the CRISP-DM methodology Dataset 1 Dataset 2 Dataset n
- 8. People and Governance ? PolyglottVS dictator Problems : • Collaboration between technical and non technical profiles inside a single project • Nécessary collaboration between business and tech teams to adress transversal projects accurately Focus : • Promote diversity • …within a workflow centric environment
- 9. End to end, from prototyping into production Do it you way …
- 10. …and scale!
- 11. DataLab Organisation Data Lab Lab Environment MultydisciplinaryTeam: Direction/ Project Management Business Analysts Data Miners / Data Scientists Production Environment Business needs Internal Data sources External datasources Missions : Priorisationof the business needs Prototyping /Agile solution engineering Support for Apps deployment Business Applications Marketing CampaignAutomation Reporting webanalytics Data as A Service Platform Conceptionof“DATAPRODUCTS” Integration of DataProducts OptimisationEngine Real Time Scoring Data Flow Insights & Services Processing chain API Deployment
- 12. Thank you !