Track A
Software engineering
You like building things and can spend an afternoon on one bug without giving up.
First experiment: Build a small inventory or to-do API with three automated tests.
Field Guides / Breaking into technology
Career guide · last checked November 19, 2026
Start where you are. Try the work before you commit. Build evidence an employer can check.
For service members in any job, veterans, students and career changers. You do not need a mentor, a technical background or a particular military specialty to use it.
Pick the line that fits you best. Everyone meets again at the foundations and the seven tracks.
Technology work is not one job, it keeps changing, and the entry level is harder than it was a few years ago.
Starting point
There is more than one good way in. The right one depends on how you learn, what you can afford, and whether serving appeals to you.
Whichever you choose, start free now: Harvard CS50x is a well-regarded free introduction to computer science.
Starting point
Your specialty does not decide your future in tech. Your timeline does. Work back from your separation date.
24+ months
12 months
6 months
Last 90 days
Starting point
You have benefits that can pay for most of this. Check each one before you choose a program.
Starting point
Your previous field is an asset, not a detour. The best roles for career changers combine what you knew with what you learn.
Everyone
These basics apply to every track. Already technical? Check yourself against the exit test and move on.
| Module | Learn and show | Free starting point |
|---|---|---|
| Computers and operating systems | Install a Linux virtual machine; troubleshoot file permissions and processes. | CompTIA A+ objectives (CompTIA) |
| Networking | IP, DNS, HTTP and routing; diagram a small lab network. | CompTIA Network+ objectives |
| Coding | Write a Python script that reads a CSV file and handles errors. | CS50 Python |
| Linux and Git | Shell, SSH, permissions, branches and pull requests. | GitHub Skills |
| Data and math | SQL joins and aggregations; algebra and basic statistics. | CS50 SQL, Khan Academy |
| AI literacy | Explain training versus use, why models are confidently wrong, and how to measure them. | Google ML Crash Course |
Exit test: explain what happens when a browser opens a website, find your way around Linux, commit a Python script to Git, run a SQL join, and explain why an AI model can be wrong. If one is shaky, repeat that module.
Everyone
Spend two to five hours on each experiment that interests you, using public data. Afterwards, score how much you enjoyed it and how much you learned from 1 to 5. Pick a primary track and one neighbor.
Track A
You like building things and can spend an afternoon on one bug without giving up.
First experiment: Build a small inventory or to-do API with three automated tests.
Track B
You like math and experiments, and you are comfortable not knowing whether something will work.
First experiment: Train a simple text classifier with scikit-learn and compare it against a naive baseline.
Track C
You like finding patterns and explaining what they mean to someone who has to make a decision.
First experiment: Load a public dataset into a database, answer three questions with SQL, and chart one answer.
Track D
You know a second language well, or you are fascinated by how meaning survives translation, and you want to work on how software handles language.
First experiment: Take 30 public sentences, compare two machine translations, and categorize the errors.
Track E
You like keeping systems running and working out why something broke.
First experiment: Run an app in Docker, break it on purpose, and fix it from the logs.
Track F
You like investigating, thinking like an attacker, and writing down exactly what happened.
First experiment: Collect logs from a machine you own, find something unusual, and write a short incident note.
Track G
You are skeptical, careful and a good writer, and you want to find out where AI systems fail and why.
First experiment: Test a public AI chatbot with 20 prompts and a scoring rubric, and record every failure.
Still unsure? Software engineering and data analytics keep the most doors open. Start one while you keep exploring.
Everyone
A certification is worth it when it closes a real gap, is required by jobs you want, or someone else pays for it. It never replaces evidence that you can do the work.
| Credential | What it covers | When it pays off |
|---|---|---|
| CompTIA A+ | Hardware, operating systems, support (two exams) | Help desk and IT support. Optional if you can already do the work. |
| CompTIA Network+ | Protocols, routing, DNS, troubleshooting | Networks, cloud and cybersecurity foundations. |
| CompTIA Security+ | Identity, risk, cryptography, threats | Common baseline for DoD cyber and IT roles (see DoD 8140 below) and for cleared jobs. |
| CompTIA Linux+ | Linux administration and scripting | Cloud, infrastructure and security roles. |
| Azure AI Fundamentals (AI-901) | AI workloads and concepts on Azure | AI literacy only (Microsoft). Not an engineering qualification. |
| A cloud associate certification | AWS or Azure architecture or administration | Cloud roles, after hands-on use. |
Check current objectives with CompTIA, Microsoft, AWS, Cisco and ISC2. Before paying, check whether Tuition Assistance, credentialing assistance, a degree program or a free program covers the exam.
Everyone
I don’t rank schools. Compare them on the same few questions.
| Degree | Best fit |
|---|---|
| Computer Science | Software, AI and NLP; the broadest default |
| Software Engineering | Applications and backend systems |
| Data Analytics, Statistics or Data Science | Analytics and evaluation; more math for data science |
| IT, network or cloud engineering | Infrastructure, operations and MLOps |
| Cybersecurity | Defense, risk and assurance |
If you are aiming at DoD roles
Only relevant for DoD cyber, IT, data and AI positions, civilian or contractor. Skip it otherwise.
DoD 8140 sets qualification requirements by work role. Basic, Intermediate and Advanced describe the role’s required proficiency, not your rank or years served. Qualification can come from education, training or a certification, and a role may add on-the-job requirements. Your workforce manager or employer confirms what a specific position requires (official 8140 documents).
Caution: the examples below follow an unofficial transcription of the 8140 matrix (immunitysec.com) because the official workbook could not be retrieved for this review. They are examples, not the full matrix, and not a compliance determination.
| Work role | Basic examples | Intermediate examples | Advanced examples |
|---|---|---|---|
| 411 Technical Support Specialist | A+, Network+, Security+ | GSEC | SSCP, SecurityX |
| 441 Network Operations Specialist | Network+, Security+, Cloud+ | SSCP, GCIH | CCNA, CCNP Security |
| 451 System Administrator | A+, Network+, Linux+ | Security+, Cloud+, SSCP | SecurityX, CCNP Security |
| 511 Cyber Defense Analyst | ISC2 CC, GFACT | Security+, PenTest+, GSEC | CySA+, GCIA |
| 521 Cyber Defense Infrastructure Support Specialist | A+, Network+ | Security+, CySA+ | GCIA, CISSP-ISSEP |
| 531 Cyber Defense Incident Responder | ISC2 CC | Security+, GCIH, PenTest+ | CySA+, GCFA |
| 612 Security Control Assessor | — | Security+, CGRC, PenTest+ | CISA, CISSP, CySA+ |
A dash means no example selected. Work role codes on the track pages come from the DoD Cyber Workforce Framework and are for orientation only; data and AI role mappings were not checked against an official matrix.
Everyone
Medians across all experience levels, not starting salaries. Use real postings and offers for the first number you expect.
| Occupation | U.S. 25th percentile | U.S. median | Utah 25th percentile | Utah median |
|---|---|---|---|---|
| Computer user support specialists | $49k | $62k | $47k | $61k |
| Network and computer systems administrators | $78k | $99k | $78k | $96k |
| Web developers | $64k | $93k | $66k | $100k |
| Software developers | $105k | $136k | $101k | $129k |
| Software QA analysts and testers | $80k | $104k | $62k | $82k |
| Information security analysts | $98k | $129k | $77k | $100k |
| Data scientists | $86k | $120k | $79k | $108k |
| Database administrators | $80k | $105k | $101k | $136k |
| Interpreters and translators | $47k | $60k | $50k | $66k |
BLS May 2025 occupational wages via O*NET (each row links to its source), rounded to $1,000. The 25th percentile means a quarter of all workers in that occupation, at every experience level, earn less; first offers are often below it. AI, NLP, MLOps and assurance jobs have no wage series of their own, so the track pages use the closest occupations. No clearance, bonus or equity premium is included. Utah figures because that is where I am; O*NET has every state.
Everyone
Use target jobs as feedback, not as a reason to collect every credential.
Everyone
Write it down, print this page if it helps, and revisit it every month.
If you have a mentor, bring finished work and your posting analysis, not a list of courses.
Written by Matthew Bonham with AI drafting assistance. Every fact on these pages links to its source, and the sources below were last checked on November 19, 2026. The learning sequence is my recommendation, not a requirement of any government agency.
No affiliation with the DoD, the VA, or any school, employer or certification provider is implied. Spot an error? Tell me.