Moving Into Data Science
If you have a budding interest in the field of data science, it’s probably good to start with what a data scientist actually is. Unfortunately, there isn’t a uniform understanding of what defines a data scientist, with debates covering this across a number of blogs, articles and books. While I can’t provide a definitive answer (there is no formal qualification that says “hey I’m a data scientist!”), the info-graphic below by Drew Conway provides a somewhat widely accepted understanding of the skills one needs to be considered a data scientist.
It’s worth noting that this framework suggests the right mix of skills for a “data scientist”, which is not the only role within the field of ‘data science’ (though generally other roles are composed of these three main skill areas in varying degrees). Some typical roles you might find within the field include data engineer, data analyst, data scientist or machine learning engineer.
What do I do at Rezdy?
I have recently joined the Insight’s team at Rezdy as an analyst. My background has been in corporate finance and management consulting, with a transition through a Master’s program to the field of data scientist. As a SaaS business providing booking software to the tours and activities sector, Rezdy holds lots of interesting information about our customers (we call them suppliers), travel agents (simply agents) and the end-user (a customer to a tour or activity provider). My time as an analyst is generally split between analysis work, stakeholder engagement and strategy.
To be a little more specific, analysis work is generally split between three main tasks:
- working on the pipeline¹;
- creating and disseminating reports; and
- performing ad-hoc or project driven analysis
On the stakeholder engagement and strategy side, my time is largely spent working with squads to define problems, sharing and discussing results of analysis and researching tools or domain insights.
What sort of skills does this require?
I caveat this with the fact that I am new to the data science generally. As an analyst, my role falls heavily down the business side of the field. My previous experience and training are far from the developed skill set of a computer scientist or mathematician. That said, you need to know enough to be deadly in order to be effective as an analyst. Daily I’ll be writing in SQL, a high-level declarative programming language for working with databases. This might seem daunting to those from consulting, finance or more business-oriented backgrounds, where Excel is king. While Excel is still used on a somewhat regular basis, it is generally for storing information quickly or sending it over to someone, and is generally unsuited for anything of permanence. If you’re considering moving into a role as an analyst, you will likely need to know some flavour of SQL (it varies under the database system you’re using).
A broader understanding of data infrastructure and programming concepts are also incredibly useful as an analyst. While you won’t need a deep knowledge of how things work, understanding computational complexity (and therefore why a query might not run in a reasonable time), or knowing how to structure data as a graph (to feed into clustering software) is extremely valuable. These were things I spent some time learning on my own and then further developed through formal education².
Any soft skills you have (communication, managing people, critical thinking, etc.) will do wonders as an analyst (and for most jobs really). A typical workflow for an analyst might consist of meeting with a stakeholder to understand their problem or need, working out how to capture and store the information necessary to solve this problem or provide insight, engaging with technical stakeholders necessary to enable the ETL (Extract Transform Load) process (if you can speak even a little engineering here, you’ll go a long way), then comes analysis techniques and preparing information for presentation, and finally conveying the information in a way that is understandable for the stakeholder. Communication is paramount in this process and any skills you’ve learnt that will aid this process will serve you well. It really can’t be stated enough how important this is to be a successful analyst.
In Conclusion
And that, at a very high level, is what a data analyst does. As you move further in your career you might look to move into a position of management or delve deeper into the technical side as a data scientist or data engineer. If this sounds like something you’re interested in, the Rezdy Insight’s team is always looking for new hires along the way. Find us at https://www.rezdy.com/careers/
Footnotes
¹ if you’re interested in the data science stack we use here at Rezdy, this https://medium.com/rezdy-engineering/an-introduction-to-data-at-rezdy-53b12d9935f5 blog post from our head of Insight’s covers it in good detail
² If you haven’t had much experience programming and want to learn some for free beforehand, this free MIT course offered by Edx https://www.edx.org/course/introduction-to-computer-science-and-programming-using-python-0 builds a really solid foundation (it also has a follow-up on some basics of data science which is also good)
