TECH CAREERS
Types of Roles
Most people in the military should look for Program Manager style roles. Potentially Strategy and Operations (StratOps). Some will be competitive for Technical Program Manager roles. Most people simply don’t have the technical experience for Data Scientist; Software Engineer; ML Engineer; or Research Scientist Roles. These are possible, but not likely if you haven’t had a very tech heavy military experience. Most people are also not competitive for Product Management Roles. Have veterans in tech companies look at your resume to calibrate on what is possible for you. Most companies have at least an informal hierarchy of roles that impact future career growth and mobility. You can read discussions on various forums about how roles compare e.g. at some companies SWEs can easily switch into CE roles, but the reverse requires a rigorous interview process.
PgM = Program manager; manages large, complex, interdependent processes at a company (e.g., runs the Hiring our Heroes program at Microsoft).
TPM = Technical Program Manager; like a PgM but needs more technical insight and works with engineers (e.g., allocates who uses which GPU cluster at Meta or helps drive execution of a product). Usually more execution than strategy
PM = Product Manager; defines vision and strategy for a product, works with engineers to build it. Balances user needs, business needs, technical capabilities.
SWE = Writes code, works with PM to execute the mission. At higher levels will be more focused on influence + system design + defining requirements
Comes in individual contributor (IC) and manager tracks. You can advance in both. But IC is probably safer
If you’re serious about being a SWE, but don’t have quite the technical experience, there are software engineering bootcamps that are full-time, many-month programs (vary in length) that are worthwhile to consider.
Data Scientist = These roles vary widely from company to company and even within companies. Pulls, cleans, analyzes data at a basic level (e.g. using SQL queries from databases). Makes dashboards and visualizations for decision- makers. Other more advanced roles involve building and training statistical or ML models (sklearn, pytorch, tensorflow). In these roles, the data scientist will build the basic model and eventually hand it over to a SWE for actual deployment. Note, you may or may not develop models (sklearn, pytorch) but you won’t be training large language models at large companies
ML Engineer = Implementing ML training and systems. Very ML heavy + SWE heavy. Not developing new algorithms but might train models. Very popular. Lots of flavors (inference, low level optimizations, distributed training). Typically, ML engineers will have a strong background in a least one application area, such as natural language processing (NLP) or computer vision (CV). If you want to work on Pixel cameras, but your background is in voice translation, you won’t get hired. So it’s best to find roles that fit with your ML background.
Data Engineer = Enabling Data Science, either for Business Analytics or Machine Learning. Range is quite large but primarily focused on getting data from a variety of systems to a centralized Data Warehouse to make Data Scientists’ lives easier. Important part of a company looking to scale their data operations.
Business Analyst = Business or Data Science background people (degrees range from STEM including Computer Science to Economics and Business) that are using data to answer business questions, usually to executives in the format of a dashboard or presentation. (see picture below to differentiate)
Analytics Engineer = Hybrid Data Engineer / Business Analyst, usually with a STEM background but understands business needs. (see picture below to differentiate)
Research Scientist = Doing research at the frontier of the field of ML, language models. Works at an extremely low level (e.g., better optimizers and activation functions), not just vanilla training of scikit-learn models. Requires a PhD in a related field, unless there are very special circumstances. Top tech companies also expect publications in top conferences and journals, and recruiters even screen by h-index - the bar is very high here.
CE = Customer Engineer = Sales-focused engineer. Building prototypes for prospects to show the value of your product before the prospect signs a sales contract. Sometimes overlaps with cloud solutions architect.
Strategy?