Whether you work in a startup, a medium company, or a large one, you probably heard about data science and the benefits you can yield from it. Improving marketing, CRM, product… so many possibilities. You probably have some data operation going on. Logging, BI, some business analytics. But the next step is somehow vague. How do you start this process of integrating data science in your organization?
Data science is a tricky. A whirlwind of material, domains, practices funnelled into a single, somehow hyped profession, that should take care all of your data, and yield competitive advantage over your clients.
Many words were written about it, but most of them address the data scientists: how to become one, algorithms explained, many nice use cases, praises and criticisms. A smaller part of write-ups address the data science process, setting, operation, which is, in fact, crucial. An even smaller part of writings actually address the business owner or manager, who have to plan, design, hire, manage and evaluate this operation.
Many questions remain unanswered: How to build a team? should it be an organic team, or split by business tasks? what knowledge should they posses?
And the workflow? should it be agile, or more research oriented? perhaps some kind of a combination of both?
And the infrastructure: every data scientist with his own python, or everyone should work on a big all inclusive framework
We, in Shibumi AI addressed these issues for dozens of times for the last 3 years, and we feel that the time has come to incorporate a series of articles, that will externalize our experience, and hopefully help others that struggle with some of these issues.
In this series I’m going to share a bit of my experience as a solution provider, to tell you about a few things that worked better and others that didn’t and perhaps give you some insights.
Planned structure of series
1. Intro (this)
3. Why do you need data science
4. How to start data science in your organization
5. How to integrate data science in your organization
6. How to Build data science infrastructure
7. How to Use external Help
8. How to plan a project
9. How to manage a project
10. How to recruit personnel
Conclusion
This is planned to be a go-to series of data science 101 for management as of current state. There was no better time to start data science than now, and hopefully this series will make it easier for you.
Comments