Hello all, I am an oracle erp developer with over 9 YOE looking to transition to DE field, was wondering if someone could share their experiences who faced this situation. My current skils include: SQL, PL/SQL, Java(Intermediate), Python(learning stage). Aware of Hadoop concepts. I have looked up some courses online viz. Data Engineering nano degree course on Udacity, not sure how much that’s hepful in transition especially have to invest $2k for that. There’s a course from insightdatascience which looks interesting but giving 7 weeks full time to them working full time in another job and being on H1B(uff) doesn’t work well. Any inputs are appreciated. #dataengineer
Thank you so much for taking time and providing your inputs.
source: 9yoe data engineering at startups, hire data engineers overall advice: first of all, expect to take an entry level position. your oracle experience isn’t going to translate directly and there’s also a cultural bias against “old school” enterprise reporting even though ultimately a lot of the concepts are similar. the cirriculum for the udacity course looks decent but IMO it’s much better to be rock solid on coding + SQL + data modeling and be familiar with the basic concepts of modern data systems than to have used a bunch of different tools in a non-production environment. having keywords like AWS/spark/airflow/etc on your resume can help you get past HR screens but ultimately skill in those areas is difficult to evaluate and so interviews will predominantly focus on coding and SQL. topics to study: sql: most important skill overall. you should know sql inside out. all the joins, window functions, case statements, how to pivot data, how to structure queries using CTEs. also have a feel for performance stuff like indexing, sorting, and partitioning (for distributed DBs). data modeling: understand basic data warehousing concepts (kimball), how to translate a business domain into table schema, normalization and its tradeoffs, building out team-specific data marts from fact and dimension tables. coding: between python and java you'll be covered for every position out there. beef up your python, especially working with dictionaries and lists, which will come up in 90% of interviews. focus more on data structures than algorithms but understand big-O especially as it relates to sorting. if you can do most leetcode easy and some leetcode mediums you’ll be fine. system design: be able to design a ETL process. understand separation of storage and compute, batch vs streaming, incrementally processing data, logging and monitoring, data lineage. know basics of how spark works but i wouldn’t bother going deep into it as it’s a huge topic and you’re not going to be able to learn it without job experience. if you have questions feel free to DM me. good luck!