I have an annual review in a little over a week and I’m feeling like my career path lacks direction.
I’ve worked at my company for 3.5 years as a Data Migration Analyst, and was promoted to a Senior Data Migration Analyst about 8 months ago. My day-to-day generally involves:
Migrating customer data to our software (working with SQL and JSON files)
Attending daily Dev-Ops meetings and doing tasks in that area (ie. shell scripting, database management) on both AWS and Azure, although we are moving exclusively to AWS shortly
Lead a team of 3 other Data Migration Analysts
Doing custom requests on customer DB’s (SQL scripting for their large updates)
Handle miscellaneous requests for other departments
I did my undergraduate degree in Data Analytics & Finance, with minors in CS and IT. I also have a Masters in Data Science.
My dilemma is that I feel that I am a master of none. I have a lot of general skills, such as SQL, Cloud Technologies and Database Management, but I’m not an expert. I also have a strong background in stats, ML and python/r programming from my undergrad/graduate degrees - all of which are not being used.
I enjoy what I do, but I want to follow a path where I’ll make more money and have hard skills that contribute to a strong resume. More importantly, I want a job that has strong prospects in the future as well.
I’m currently trying to weigh my options:
Deep dive into cloud technologies and become an expert in cloud engineering or something along those lines
Improve my python programming skills and focus in data engineering
Try to get back to my roots and find work in DA/DS/BI since it’s the bulk of what I studied
Shift to something less technically. Become a migration specialist that covers more of the process than just the data. 5. Keep your current job and shift your life focus to something other than your career.
Really I’m just trying to help you brainstorm here. I think any of the options could be right depending on you.
I completed step two. I was a data scientist once, but because I like developing things and seeing my work truly benefit the organization, I shifted more into data engineering job. I was sick and tired of creating models and undertaking analyses that would never be used.
I am currently employed as a data engineer, and things are going great.
Are you me? I got tired of having to build nonsense or to answer unanswerable questions. DE actually, consistently delivers value. If I found a cool DS project I might jump over but almost all DS relies heavily on engineering that most DSs don’t do well.
There are a lot of us. I’ve heard similar stories quite a bit over the years. Some days I miss the more analytical work, but I agree – oftentimes the modeling part of the job is like 10% and the infrastructure around that model is 90%.
I get to do a bit of machine learning work as a DE but definitely not a ton.
As DS (lead) you need to do a lot of stakeholder management to align deliverables and how they play into management. You need to follow KISS often, when you actually would go full scientist, just to serve a audience that has neither the skills nor the time that would be appropriate to explain the problem, let alone a reasonable approach. You just need to pitch the solution in a slide deck. Spent more time on optimizing slidedcks than actual experimental designs.
As DS (ML researcher) you need to get into quite a lot of new and trendy subjects that are not easy to read up. I am in for every bigger method since SVM, and it becomes a little tiresome. Especially when stuff becomes super complicated and you need to understand how it interplays with cloud infrastructure as well (ML engineer). Thank god the topic is now broader than ever, so at least documentation is becoming much better…
As DS (operational) you need to get often so deep into the systems, understanding the process and data to a degree where you question a lot of the context. Additionally you are probably working on a challenging DS implementation, too. So when i come home, it takes me some time to disengage from the complex DS thoughts and shitty processes I saw.
As DS (programmer) you have these forsaken PoC implementations that you will never come back to, clean up or optimize. Its a bit sad, and leads to an awkward Git full of badly documented unproductive code…
As DE all these problems are gone (or at least less pronounced).
But also as DE i felt i can do even better for my needs. Now I am more into the project evaluation and planning. Well… So basically now I am only doing DS as a “black box” theory.
(or at least less pronounced).