Engineering vacancies name a stack and a scale. Screening matches the exact technologies in the posting, then a human looks for evidence you ran them in production: requests, latency, coverage, incidents. A resume that lists twenty technologies but no numbers loses both passes.
Keywords that appear in Software Engineer job descriptions
- Python
- Java
- TypeScript
- React
- Node.js
- REST APIs
- microservices
- SQL
- PostgreSQL
- Git
- CI/CD
- Docker
- Kubernetes
- AWS
- unit testing
Use only the keywords that are true for your experience — screening software matches them, but humans read them.
How to rewrite your bullets for a Software Engineer job
Developed backend features for the platform.
Built and shipped 12 REST endpoints in a Python/FastAPI service handling 2M requests/day; cut p95 latency from 480ms to 210ms.
Fixed bugs and improved code quality.
Raised unit-test coverage from 40% to 75% on the payments module and cut production incidents by a third.
The numbers above are placeholders — swap in your real ones.
Common Software Engineer resume mistakes
- A wall of every technology ever touched: screeners match the vacancy's stack — the other fifteen names are noise that dilutes the signal.
- No scale anywhere: requests per day, users, latency, coverage. Without numbers, 'built services' could mean anything.
- Describing the team's work instead of your commits — 'the team migrated to Kubernetes' says nothing about you.
Summary line patterns
- Backend engineer (Python, PostgreSQL, AWS), 5 years — APIs, performance work, and CI/CD in product teams.
- Full-stack developer: React + Node, 3 years in startups, shipping to production weekly.
Knockout requirements to check before applying
- Work authorization
- The vacancy's stack listed as required
- Seniority / years of experience
- Degree requirements at some companies