DataTalk #42 What is DataOps? Becoming a truly data-driven organisation

En 2017, Gartner annonçait que 85% des projets (big) data échouaient à se déployer en production. En 2019, ils confirment la tendance en annonçant que seulement 20% des observations analytiques de pointes aboutiront à un résultat commercial. La faute à une absence, dans le cycle de travail en Data Science, de cadres organisationnel et technologique cruciaux aux nouvelles exigences en termes de réduction du “time to market”.
Durant cette présentation, je souhaite vous présenter la discipline qui vous aidera à rassembler vos donnée, votre équipe, vos outils et vos process afin de devenir une véritable organisation data-driven et ainsi faire mentir les chiffres de Gartner. Cette discipline est la DataOps.

Transcript

1. What is DataOps?
   Samia Drappeau
   Becoming a truly data-driven organisation
   Toulouse 2020
   

   View Slide

2. About me
   2
   • Astrophysicist for 8 years
   • Data Scientist at Continental for 2
   years
   • Now Fullstack Data Consultant and
   DataOps enthusiast
   —
   • Toulouse Data Science meetup host
   • OpenClassroom mentor
   

   View Slide

3. Data (Integrity|Quality|Reliability)
   3
   

   View Slide

4. What is DataOps?
   4
   DataOps is an automated, process-oriented methodology, used by
   analytics and data teams, to improve the quality and reduce the cycle
   time of data analytics.”
   ~ DataOps Manifesto
   

   View Slide

5. What is DataOps?
   4
   DataOps is an automated, process-oriented methodology, used by
   analytics and data teams, to improve the quality and reduce the cycle
   time of data analytics.”
   ~ DataOps Manifesto
   

   View Slide

6. What is DataOps?
   5
   Culture
   Process
   Tools
   

   View Slide

7. Data-Driven Company Culture
   6
   What is DataOps?
   Culture
   Forbes Article: The four key pillars to fostering a data-driven culture
   

   View Slide

8. 7
   What is DataOps?
   Processes Process
   

   View Slide

9. Tools
   8
   What is DataOps?
   Tools
   

   View Slide

10. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

11. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

12. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

13. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

14. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

15. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

16. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

17. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

18. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

19. How to implement DataOps?
    10
    The Seven Steps of DataOps
    

    View Slide

20. 11
    How to implement DataOps?
    Technology Ecosystem
    

    View Slide

21. Last words
    12
    DataOps can help you bring together your data, your team, your
    tools, and your processes to become a truly data-driven
    organisation.
    

    View Slide

22. ANNEXES
    

    View Slide

23. Gartner Hype Cycle for Data Management, 2018
    14
    

    View Slide

24. Adding the Ops in Data
    

    View Slide

25. Sources
    16
    ▸ https://www.valdas.blog/2019/04/17/data-ops/
    ▸ https://www.forbes.com/sites/brentdykes/2019/03/28/the-four-key-pillars-to-
    fostering-a-data-driven-culture/#61aea7a57d90
    ▸ https://medium.com/data-ops/dataops-is-not-just-devops-for-data-6e03083157b7
    ▸ https://www.experis-it.fr/dataops-vers-une-entreprise-data-driven/
    ▸ https://www.datakitchen.io/blog/dataops/2019/07/03/DataOps-Data-
    Architecture.html
    ▸ https://www.gartner.com/en/newsroom/press-releases/2018-09-11-gartner-hype-
    cycle-for-data-management-positions-three-technologies-in-the-innovation-trigger-
    phase-in-2018
    ▸ https://www.dataops.dev/
    

    View Slide