Resume
lastmod 2026-07-19created 2025-07-215 edits
A résumé (Korean iryeokseo, 履歷書; North Korean Munhwaŏ: riryeokseo) is a document submitted to a company or other organization to get a job interview, which summarizes or lists in chronological order a person's personal details, education, work experience, and so on. In English and French it is called a résumé, and it is also called a CV, short for curriculum vitae. A résumé is shorter and simpler than a curriculum vitae.
Résumé formats
Quantitative metrics for programming résumés
A page organizing, by category, quantitative (numeric) metrics you can use in a programming résumé.
Performance improvements
- Reduced response time/latency: cut API response time by 85%, from 800ms → 120ms
- Increased throughput: tripled requests processed per second
- Reduced build/deployment time: CI/CD pipeline run time 15 min → 4 min
- Optimized memory/CPU usage: reduced memory usage by 40%
Size/scale
- Data volume processed: built a pipeline processing 5 million logs per day
- Concurrent users/traffic: designed an architecture supporting 10,000 concurrent users
- Codebase size: a system of 12 microservices and 200,000 lines of code
- Cluster/node size: managed resources for an 8-node k3s cluster with 4 GPUs
Cost reduction
- Infrastructure cost: cut monthly cloud costs by 30% (resource optimization)
- Labor/time saved through automation: saved 10 hours per week by automating the manual deployment process
Stability/reliability
- Uptime: achieved 99.9% service availability
- Fewer incidents: production incidents 5 → 1 per month
- Mean time to recovery (MTTR): average recovery time 40 min → 8 min
- Test coverage: unit test coverage 45% → 85%
Business impact
- Conversion rate/revenue contribution: conversion rate up 12% from checkout flow improvements
- User growth: active users up 25% after a new feature launch
- Retention: 7-day retention up 18%p from onboarding improvements
Productivity/collaboration
- Code review/PR handling: reviewed an average of 40 PRs per month, average 1.5 days to merge
- Team size/leadership experience: led a team of 5 developers, managed sprints for 8 features
- Documentation/onboarding: onboarding documentation cut new-hire ramp-up from 3 weeks → 1 week
ML/AI-specific
- Model performance: improved mIoU from 62% → 78% (handling class imbalance)
- Training time: training time 12 hours → 3 hours by adopting distributed training
- Inference speed: reduced inference latency by 55% with TensorRT optimization
Writing tips
- The Before → After format is the most persuasive (e.g., 800ms → 120ms)
- If you lack exact figures, be ready to justify even your estimates (prepare for verification questions in interviews)
- List only what you personally measured/contributed; distinguish team-wide results from your individual contribution
See also
Favorite site
Résumé static sites
Articles
- If you're senior, own a graph | GeekNews
- The more senior an engineer, PM, or designer is, the more they should have at least one graph of a key metric they own, and this is one of the fastest ways to do good work
- Most of the important problems to tackle at the senior level can be expressed as graphs, and you are expected to directly own metrics that show change over the long term, such as page count, performance, cost, revenue, and churn rate
- A graph reveals scale, trend, volatility, and timing all at once better than a sentence like "reduced it by 15%", and is the most powerful communication tool for explaining impact and earning verifiable trust
- A loop that combines graphs with goal setting and check-ins makes performance tracking and feedback extremely simple, and can lead to a difference on the scale of high performance versus losing your job
- A good graph should be one of a few important ones that others refer to often, and the core role of a senior is to own a single graph in areas such as quality, support, revenue, or retention all the way through and make it move