In the Last few weeks, I have received several messages, relating to starting a career in Databases and Analytics. Most of the questions revolved around the below and I have tried to answer all of them here (You can also Join my list to get latest updates on articles and courses)

Common questions:

1. I want to have a career in Databases, where do I Start?
2. What exactly does a BI engineer do?
3. Is Data engineer different from DBA?
4. What is data science?
5. How do I become an Analyst?
6. What kind of Salary can I expect in the US as a data engineer?
7. What kind of projects does a Data Engineer work on?
8. I want to Switch Careers. Where do I start?
9. I am not good at Coding, but interested in databases. Where should I start?
10. How do i start learning Data-Science.
11. If I learn SQL server, How do i keep up with other technologies? Oracle, Hadoop, Teradata etc.?


First of all, let me tell you that the field of analytics is an interesting one. If you are new to Analytics or considering a switch of careers, first and foremost I would recommend is to be comfortable writing basic SQL queries. This forms the fundamental part of your work, even if your work doesn't involve active coding. It will give you immense freedom and will allow you to do your job without depending on anyone.

If you are absolutely new, you should start with 5 simple concepts:

1. Creating your first SQL table and storing some data.
2. Writing some basic 'Select' queries and understanding how to filter data (where clauses).
3. How joins work and what are the different types of joins?
4. What is aggregation and what are the various types of aggregations?
5. Getting acquainted to some basic functions like Date, String etc.

I would recommend taking my course on "Introduction to SQL Querying".

Once you are comfortable with the SQL ecosystem and querying, i would recommend you to take it a bit further and learn some advanced querying.
1. Views, Triggers, Computed columns.
2. Writing Stored Procedures, User Defined Functions.
3. Getting a grasp on how transactions work.
4. How to write Dynamic queries (this is very useful and interesting).
5. Learning some interesting Datatypes - Like XML, Spatial Etc. (Though you might not use them in Day to Day projects, its interesting to learn them)

If you are interested, you can checkout my advanced course here

Once your fundamentals are solid, you can choose your career path.
As a data-engineer, mostly you will be expected to architect and build Data marts so that other teams can use it for effective reporting. It is very essential that you know the business domain thoroughly. This will help you step up as a consultant and a data engineer, where you will be in a position to dictate "correct" and "incorrect" ways of building marts. Make sure you are rock solid on Database architecture, Querying, a little DBA stuff and your business domain. A second path as a data engineer will be an ETL developer. An ETL developer will be responsible to move data from one system to another. By moving, what I mean is collecting data from say Amazon S3 or other Amazon Databases or CSV, or Oracle and stitching/cleaning them and storing them in say SQL server in a report able format (data marts). This will require you to know more about ETL and tools.

If you are interested, you can refer to my course "Learn ETL using SSIS"


Another interesting path is a Reporting Analyst. Every company will have different expectations from an analyst. Generally reporting analysts are not expected to start from scratch (like building Tables etc.). The data will already be provided to you. You need to tell a good story and recommendations with that data. You will generally be expected to know a few reporting tools like tableau or QlikView or even Excel. You should be comfortable enough to Quantify a problem, how big is it and is it worth yours and your team's time to analyze it. You will need to study data after various A/B tests and provide your analysis. Many critical decisions will be taken based on your suggestions. You should be comfortable identifying outliers in a data. This is an interesting field. You might meet a lot of folks, get to know businesses, and consult on need basis.
Due to popular demand, I have started to record some courses on Reporting. You may Subscribe yourself here to keep updated on latest courses.

A third path is the Data Science Path. To be honest, there will be a little learning curve if you are planning to switch careers and have no experience at all. If I were to give you a 500,000 ft view of what is data science and Machine Learning, I would say, it is identifying some data, inputting the data to an Algorithm, and the algorithm predicting something. The challenge here is 1 - Choosing the right data as input. This is where you will spend a lot of time. You will need to know the domain well, to choose, clean and modify the data before supplying to the algorithm.
2 - There are plenty of algorithms out there to achieve various results. Algorithms for predictions, Image Classifications, Videos, Clustering etc.
You need to know which algorithm to choose and best use for your situation. For example - You must have heard of Regression.
To give an example - it may happen that you give the same set of data to 2 different engineers. Both may use the same algorithm, but one of the engineer may produce a better accurate prediction. Why? He may have used different techniques to clean the data. This is where simple business domain knowledge and Mathematics help.
If time permits, I will definitely record a free course on Data science algorithms.

If you have more questions - feel free to send me a note.

Thanks,

Rakesh

http://www.rakeshgopal.com