Software Engineer
Designs and builds the software everything else now runs on. Very well paid, genuinely enterable without a degree — and harder to break into than it was five years ago.
- Work environment
- office, remote-capable
- Typical hours
- moderate with peaks
- Stress
- moderate
- People contact
- small team
- Income
- very strong
- Degree needed
- No — there is a non-degree routeO*NET 2026O*NET 2026From the O*NET occupational database.O*NET 30.3 — Software Developers (15-1252.00): tasks, job zone, education distribution
Stress. Usually low-grade and continuous rather than acute: deadlines, on-call rotations, and the particular frustration of a bug you cannot reproduce. Nobody dies if you are wrong, which changes the character of the pressure completely compared with clinical or safety-critical work.
Hours. Broadly predictable and among the most flexible of any well-paid career, with genuine remote work available. Peaks come from launches, incidents and on-call rotations, and the culture varies enormously between a bank and a startup.
People. Far more collaborative than the stereotype suggests — most of the job is agreeing what to build with other people. But it is a small, consistent team rather than the public, and there are long stretches of uninterrupted solo work.
Income. Among the highest-paying careers reachable with a four-year degree and no postgraduate study. The variance is extreme, though — the same title pays several times more at a large technology company than at a typical employer.
What they actually do
The real tasks, not job-description language.
- Work out what a piece of software actually needs to do, and whether it can be built in the time and budget available.
- Design how a system will be structured before writing any of it — the decisions that are expensive to reverse later.
- Write code, which is a smaller share of the working week than almost anyone expects.
- Read other people's code, which is a much larger share than anyone expects.
- Modify existing software to fix defects, adapt it to new requirements, and stop it degrading.
- Build and run testing and validation, because the cost of a defect rises sharply the later it is found.
- Review colleagues' work and have your own reviewed, which is how standards and knowledge actually spread.
- Investigate production incidents — the system is broken, users are affected, and you have to find out why.
A day in the life
Examples, not measurements. Real days vary; these are what people describe as typical.
An ordinary dayO*NETO*NETFrom the O*NET occupational database.
- 09:15Catch up on messages and code review requests before starting anything of your own.
- 09:45Stand-up. Fifteen minutes on what everyone is doing and what is blocked.
- 10:00The good part: two or three hours of uninterrupted work on a real problem.
- 13:00Lunch, usually away from the screen.
- 14:00Design discussion about something being built next month. Mostly arguing about trade-offs.
- 15:00Review a colleague's changes properly, which takes longer than writing them would have.
- 16:00Back to your own work, now fragmented by afternoon interruptions.
- 17:45Push what you have, write down where you got to, stop.
A production incidentreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 5 independent accounts.Incident response is consistently described as the most intense part of the job: unplanned, time-pressured, collaborative, and followed by a blameless review process focused on the system rather than the individual.
- 02:40Paged. Something is down and you are on call.
- 02:55Dashboards and logs. Establish what broke, when, and how many people it affects.
- 03:30Mitigate first, understand later — roll back or fail over to stop the bleeding.
- 04:15Service restored. Write up what happened while it is fresh.
- 11:00Incident review. The good ones examine the system that allowed it, not the person who tripped it.
First months in the jobreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 5 independent accounts.New engineers consistently describe the first months as dominated by reading existing code, small supervised changes, and the discovery that university coursework poorly resembles working on a large existing system.
- 09:30Read code. Enormous amounts of it, most of which you do not yet understand.
- 11:00A deliberately small first task, so you learn the process rather than the problem.
- 14:00Ask a question you are worried is stupid. It is not — everyone asked it.
- 16:00Your change gets reviewed and comes back with twenty comments. This is normal and is how you learn.
Education pathway
What it actually takes, with realistic time at each stage.
Saudi ArabiaSchool to independent practice: 4–5 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 3 independent accounts.Degree length only — there is no licensing or postgraduate requirement in this field.
- Secondary school, science track (علمي) preferred3 yearsestimatedestimatedInferred by reasoning, not measured. The basis is given below.Standard Saudi secondary structure; the science track is the usual prerequisite for computer science and engineering admission, though non-degree routes remain open from either track.
- BSc Computer Science or Software Engineering4–5 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 3 independent accounts.Saudi computing degrees are consistently described as four to five years including a preparatory year at many universities.
- Working engineer0 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 2 independent accounts.No licence or registration is required to practise; employment follows directly from graduation or from demonstrated skill.
Licensing
None. This is one of the few well-paid technical careers with no licence, no registration and no professional body gate — what you can demonstrate matters more than what you hold.
Notes
Because there is no licensing, the non-degree route is genuinely open here in a way it is not in engineering or health. A strong portfolio and demonstrable skill can substitute for the qualification, though it is harder than it was.
United States — via a degreeSchool to independent practice: 4 yearsO*NET 2026O*NET 2026From the O*NET occupational database.O*NET 30.3 — Software Developers (15-1252.00): tasks, job zone, education distribution
- High school with maths4 yearsestimatedestimatedInferred by reasoning, not measured. The basis is given below.Standard US secondary structure; mathematics is the subject that matters for computing admission.
- Bachelor's in computer science or a related field4 yearsO*NET 2026O*NET 2026From the O*NET occupational database.O*NET 30.3 — Software Developers (15-1252.00): tasks, job zone, education distribution
85% of working software developers hold a bachelor's — but 5% hold only an associate degree, and the field remains formally open without one.
- Working engineer0 yearsBLS 2025BLS 2025From US Bureau of Labor Statistics wage statistics.Occupational Employment and Wage Statistics, May 2025 — Software Developers (15-1252)
Licensing
None.
Without a degreeSchool to independent practice: 2–5 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 4 independent accounts.Self-directed learning plus the time to a first role, consistently described across non-degree entry accounts. Highly variable, and the variance is larger than for the degree route.
- High school, with maths taken as far as you can4 yearsestimatedestimatedInferred by reasoning, not measured. The basis is given below.Standard US secondary structure. Maths matters here even without a degree, because technical interviews test reasoning that school maths builds.
- Self-directed learning or an intensive programme, building real projects1–3 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 5 independent accounts.Self-taught and bootcamp entrants consistently describe one to three years of building demonstrable projects before landing a first role, with the portfolio rather than the course certificate doing the work.
What gets you hired is a portfolio of things that actually run, not the certificate. Nobody checks the certificate.
- First role, often junior or contract, then compounding experience1–2 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 4 independent accounts.Non-degree entrants consistently describe the first role as the hardest step by a wide margin, after which hiring proceeds on experience and the absence of a degree stops being raised.
- Working engineer0 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 4 independent accounts.Once a first role and one to two years of experience exist, non-degree engineers consistently report being assessed on the same basis as graduates.
Licensing
None — which is precisely why this route exists at all. No employer can be required to see a qualification, because there is no qualification to require.
Notes
This route is real and people take it every year, but be clear-eyed: it is currently harder than the degree route, not easier. The first role is the entire difficulty, and the contraction in entry-level hiring has hit candidates without degrees hardest. It works best if you can build something genuinely impressive and show it, and it works badly as a way of avoiding study — the studying still happens, it is just unsupervised.
United KingdomSchool to independent practice: 3–4 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 3 independent accounts.Degree length only, with no licensing stage.
- A-levels, usually including maths2 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 3 independent accounts.UK computer science degrees consistently require mathematics at A-level, more often than they require computing itself.
- BSc Computer Science, or a degree apprenticeship3–4 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 3 independent accounts.Three-year degrees are standard; degree apprenticeships run four years and are salaried throughout, which is a materially different financial proposition.
- Working engineer0 yearsreportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 2 independent accounts.No registration required; employment follows graduation.
Licensing
None.
Notes
Degree apprenticeships are worth serious attention: you are paid throughout, the employer covers tuition, and you finish with four years of real experience rather than none. Places are very competitive.
What to study now
Subject choices made at fifteen or sixteen decide what is still possible at eighteen.
Saudi curriculum track
The science track is the standard route into Saudi computing degrees. Uniquely among well-paid technical fields, though, the administrative track does not close software permanently — because there is no licensing body, a demonstrable portfolio can still get you hired. That is a real escape hatch, but it is a harder road than the direct one.
Doors that close without these
- Without maths at A-level or IB Higher Level, most competitive computer science degrees are closed. This is the single most common mistake — students take computing and drop maths, which is exactly backwards.
- Nothing closes the field permanently, because there is no licence. But the degree route closing makes everything afterwards considerably harder.
A-Level
- MathematicsrequiredRequired by most computer science degrees — more often than computing itself is. This surprises people.
- Computer ScienceusefulHelpful but rarely required. Universities teach programming from scratch and often prefer maths.
- Physicsuseful
- Further MathematicsusefulGenuinely helps at the more theoretical universities.
IB
- Mathematics (Higher Level)required
- Computer Scienceuseful
- Physicsuseful
Degrees that lead here
The whole route on one page →- Computer ScienceThe largest destination by a very wide margin.
- Electrical and Electronic EngineeringEmbedded and systems software; electrical graduates move into software far more easily than the reverse.
- MathematicsEmployers treat a mathematics degree as evidence of the reasoning ability programming requires.
- PhysicsA very common destination, and employers read the degree as evidence of reasoning ability.
If any of those systems is unfamiliar — or you have not chosen between them yet — the exams and qualifications section covers what each one is, which subject inside it opens which degree, and when to sit what.
Getting in: how competitive
Students consistently underestimate this part.
The honest answer has changed and most careers advice has not caught up. Getting onto a degree is moderately competitive. Getting the first job is now substantially harder than it was five years ago — entry-level hiring contracted sharply while the number of graduates and bootcamp leavers kept rising. Experienced engineers remain in demand; juniors are competing hard. Plan for the first job to take months rather than weeks.
What selectors actually weigh
Grades matter for university admission and then stop mattering almost immediately — hiring is dominated by technical interview performance and by what you can demonstrate you have built. A portfolio of real work outperforms a better degree classification, which is unusual among professional careers.reportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 5 independent accounts.Consistently described across software hiring accounts: after the first role, academic record is rarely examined, while demonstrable projects and interview performance dominate selection.
Exams in the way
- Technical interviews, including live problem-solving
- Take-home coding exercises
- System design interviews at more senior levels
How many attempts is normal
Applying to dozens of positions for a first role is now normal rather than a sign of failure. This is a market condition, not a verdict on you.
Reality check
Both columns are required. A career page with no difficult parts is an advert.
The good
- The pay is exceptional for four years of study and no postgraduate requirement or licence.
- Remote and flexible work is genuinely available here in a way it is not in most well-paid careers.
- There is no licensing body, so the field is formally open to anyone who can demonstrate the skill — one of very few professions where that is literally true.
- The work is intellectually satisfying in a specific way: problems have solutions, feedback is immediate, and you can see the thing you built working.
- The skills transfer across essentially every industry, which makes it unusually easy to change sector without changing career.
The difficult parts
- The entry-level market is genuinely difficult right now, and anyone telling a sixteen-year-old this is an easy well-paid career is describing 2019 rather than today.
- You have to keep learning permanently. The specific technologies you learn at university will be partly obsolete within a decade, and this is tiring in a way people underestimate.
- A large share of the job is dealing with other people's old code and other people's decisions, not building new things.
- Sitting still for eight hours has real physical costs, and they arrive earlier than people expect.
- The extreme salary figures circulating online come from a small number of large employers and are not representative of the field.
- On-call rotations mean the job can wake you up, and incident work is genuinely stressful while it lasts.
Who this suits
This suits you if
- You enjoy the specific frustration of a problem that resists you, and the specific satisfaction when it gives.
- You can hold a complicated structure in your head and keep working on it.
- You are willing to keep learning indefinitely rather than qualifying once.
- You are comfortable being wrong in public — code review is relentless feedback.
- You want strong pay without a decade of training or a licence.
Think twice if
- You want to be moving around and away from a screen.
- You are choosing it purely for the salary figures you have seen online, which are not typical.
- You dislike ambiguity — requirements are frequently vague and change after you have built to them.
- You expect the first job to arrive quickly, because at the moment it usually does not.
- You want a career where qualifying once means you are done learning.
Salary
Ranges, not a single figure. The median matters more than the ceiling.
United States · USD per year
- Entry
- $82,460–105,210BLS 2025BLS 2025From US Bureau of Labor Statistics wage statistics.Occupational Employment and Wage Statistics, May 2025 — Software Developers (15-1252)
- Mid-career
- $105,210–171,980BLS 2025BLS 2025From US Bureau of Labor Statistics wage statistics.Occupational Employment and Wage Statistics, May 2025 — Software Developers (15-1252)
- Senior
- $171,980–214,670BLS 2025BLS 2025From US Bureau of Labor Statistics wage statistics.Occupational Employment and Wage Statistics, May 2025 — Software Developers (15-1252)
What drives the spread
Percentile bands across 1,687,890 developers at one moment, not a career track. Median was $135,980. Employer matters more than seniority here than in almost any other field — the same work at a large technology company versus a mid-sized non-tech employer can differ by a factor of two or more. These figures also exclude equity, which at the top end is a large part of total compensation and is not captured in wage statistics at all.
How pay is structured
Salary plus, at many technology companies, equity that can exceed the salary. Equity is a genuine part of the package and a genuine risk — it is worth what the company turns out to be worth.
Saudi Arabia · SAR per year
- Entry
- SAR 120,000–200,000estimatedestimatedInferred by reasoning, not measured. The basis is given below.Inferred from Gulf technology sector pay levels and the strong demand created by digital transformation programmes. No published Saudi occupational wage statistic for software roles was obtainable.
- Mid-career
- SAR 200,000–380,000estimatedestimatedInferred by reasoning, not measured. The basis is given below.Inferred from mid-career technology pay in comparable Gulf markets, where scarcity of experienced local engineers pushes rates above what the cost of living alone would suggest.
- Senior
- SAR 350,000–700,000estimatedestimatedInferred by reasoning, not measured. The basis is given below.Inferred from senior and lead engineering roles at major Saudi employers and giga-project technology functions. The upper end reflects scarce senior talent rather than typical outcomes.
What drives the spread
Estimated rather than measured, and the spread is wide. Saudi technology pay is being pushed upward by demand from banks, telecoms and the giga-projects competing for a small pool of experienced engineers. Salaries for juniors are far less inflated than for seniors.
How pay is structured
Increasingly package-based with allowances, and equity is rare outside startups. Multinational employers and the sovereign-backed technology programmes pay differently from traditional local firms.
The Saudi picture
Specific to Saudi Arabia, shown whichever country is selected above.
Does this field actually hire here
Genuinely strong and genuinely growing, and this is one of the fields where the Vision 2030 narrative and the actual hiring line up rather than diverge. Banks, telecoms, government digital programmes and the giga-projects are all recruiting engineers simultaneously, and the pool of experienced Saudi engineers is small relative to that demand. The caveat is the same as everywhere: demand concentrates on experienced people, and graduates still have to compete for a first role.
Government vs private
Government digital transformation programmes and sovereign-backed entities have become major technology employers and pay competitively, which is a change from a decade ago. Traditional private firms pay less than the sovereign-backed programmes and multinationals. Startup equity is rare and should be treated sceptically.
Saudization
Technology roles carry substantial Saudization pressure, and the shortage of qualified nationals means employers compete for Saudi engineers rather than merely meeting a quota. This is a genuine and significant hiring advantage.
Licensing and foreign degrees
None — no registration, no professional body, no recognition process for foreign degrees. That makes this one of the few fields where studying abroad carries no licensing risk on return, and one of the few where a non-traditional route into the work is genuinely viable.
Vision 2030
Among the more credible parts of the programme. Digital infrastructure and government technology are funded, staffed and already delivering, rather than announced and pending. Even so, treat any specific project's headcount promises with more caution than the sector-level trend.
Provenance for this sectionestimatedestimatedInferred by reasoning, not measured. The basis is given below.Reasoned from the absence of licensing requirements, published digital-transformation policy direction, and the structural scarcity of experienced engineers in the region. No occupational employment or wage statistic for Saudi software roles was obtainable.
Career progression
A realistic ladder, with the years each rung usually takes.
- Junior engineeryears 0–2reportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 5 independent accounts.Consistently described as a period of supervised work on well-defined tasks, with the main goal being to learn the codebase and the process rather than to deliver independently.
- Engineeryears 2–5reportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 4 independent accounts.Independent delivery of features and ownership of components consistently described as arriving in this window.
- Senior engineeryears 5–10reportedreportedConsistently reported across multiple independent credible accounts. Not a measured statistic.Based on 4 independent accounts.Senior level consistently described as marked by designing systems and mentoring others rather than by writing more code.
Where most engineers settle, and a genuinely good place to stay.
- Staff engineer or engineering manageryears 8–18estimatedestimatedInferred by reasoning, not measured. The basis is given below.Inferred from the standard two-track structure of software careers, where the technical and management tracks diverge after senior level rather than at a fixed point.
The fork: stay technical, or move into managing people. They are different jobs and many people discover too late which one they wanted.
Specialisations
One job title can contain very different lives.
- Backend and distributed systems
- Servers, data and scale. The largest share of jobs and generally the best paid.
- Frontend and interface
- What users actually touch. Overlaps heavily with design and is often undervalued relative to its difficulty.
- Mobile
- iOS and Android. A narrower but consistently well-paid niche.
- Infrastructure and reliability
- Keeping systems running. On-call is central, and pay reflects that.
- Embedded and systems programming
- Software for physical devices, where constraints are real and mistakes are expensive to fix.
- Security engineering
- Overlaps substantially with cybersecurity and pays a premium.
How this field is changing
You enter this workforce in five to twelve years, not today.
Demand: flatestimatedestimatedInferred by reasoning, not measured. The basis is given below.Inferred from the occupation's large recorded employment combined with consistently reported contraction in entry-level hiring. No published occupational projection was obtainable, so the direction is reasoned rather than measured, and it is the claim on this page most likely to change.
The occupation remains very large at 1,687,890 people in the US, but entry-level hiring contracted sharply after a period of rapid expansion, and the graduate supply did not contract with it. Demand for experienced engineers has held up considerably better than demand for juniors — which is the opposite of how this field is usually described to students.
What automation actually changes
This is the field where AI coding tools have changed the daily task mix most visibly, and being honest about it matters more here than anywhere else on this site. Writing routine code is substantially faster than it was. What has not been automated is deciding what to build, judging whether a proposed solution is actually correct, and holding responsibility for a system in production — and reviewing machine-generated code is now itself a large part of the job. The plausible near-term picture is fewer people needed for the same output at the junior end, and continued demand for people who can judge whether the output is right. That is a real reason for caution about the entry market, and not a reason to conclude the occupation is ending.
Are requirements drifting
Moving in the opposite direction from most fields. A degree is still the standard route, but portfolio and demonstrated skill carry more weight than in any comparably paid profession, and there is no licence anyone can require you to hold.
How much has really changed
The field has been repeatedly reshaped — mainframes, web, mobile, cloud, and now AI tooling — roughly once a decade, and each time the work changed substantially while the occupation grew. That is a field with a high rate of change, which suits some people and exhausts others. It is worth knowing which you are before committing.
Sideways from here
The most useful direction on this site. Going deeper only tells you that medicine contains cardiology.
If you like this, consider
- Data EngineerThe same building instinct applied to data pipelines. Less glamorous, consistently easier to get hired into.
- Security Operations AnalystThe same systems knowledge used defensively, with a clearer certification route in.
- Machine Learning EngineerAdjacent and overlapping, with more mathematics and a narrower job market.
- Electrical EngineerFor people drawn to computing but wanting physical hardware rather than pure software.
Same interest, different trade-off
Careers driven by what draws you here, with a materially different length, cost or lifestyle attached.
- Data EngineerSame building work, different material.Comparable pay and a noticeably less competitive entry market, at the cost of a lower public profile and work that is more plumbing than invention.
- Electrical EngineerThe same interest in how systems work, in hardware.A licensed, more traditional profession with slower salary growth and far more stability. You gain a career that is not reshaped every decade, and lose the flexibility and remote work.
Where Computer Science (BSc) can take you
The same degree, other destinations. Choosing this subject does not commit you to this job.
What next
Sources for this page
Last researched 2026-08-16. Every figure above carries the label of where it came from — hover or tap one to see which.
- O*NETO*NET 30.3 — Software Developers (15-1252.00): tasks, job zone, education distributionaccessed 2026-08-16
- BLSOccupational Employment and Wage Statistics, May 2025 — Software Developers (15-1252)accessed 2026-08-16
- reportedConsistently described across software engineering practitioner and hiring accountsaccessed 2026-08-16