DevOps Engineer résumé example
A DevOps or platform engineer is hired to make other engineers faster and the system harder to break, so the résumé has to show both: deploy frequency, lead time, incident counts and cloud cost on one side; the teams and services served on the other. This example gives every bullet a before-and-after number and keeps tooling in the side column, where it is scanned and not read.
Summary
Platform engineer with nine years across infrastructure, CI/CD and reliability, the last four running the platform for 14 product teams at an energy marketplace. Took deploys from twice a week to 60 a day, cut mean time to recovery from 3 hours to 25 minutes and reduced AWS spend by 38% in one year while traffic doubled. Writes Terraform and Go, runs on-call, and prefers deleting a service to adding one.
Experience
- Own the Kubernetes platform (EKS, 6 clusters, 420 services) used by 14 product teams; deploy frequency rose from 2 a week to 60 a day after moving every team to GitOps with Argo CD.
- Cut mean time to recovery from 3 hours to 25 minutes by introducing SLOs, burn-rate alerts and a one-command rollback; Sev-1 incidents fell from 11 to 3 a year.
- Reduced AWS spend by 38% (€1.9M a year) through Graviton migration, spot node pools for batch work and a weekly cost review with each team, while request volume doubled.
- Built the internal developer portal (Backstage) with golden-path templates; a new service goes from repository to production in 40 minutes instead of a week of tickets.
- Led the SOC 2 infrastructure workstream: policy-as-code with OPA, image signing with cosign and automated evidence collection; passed the audit with no infrastructure findings.
- Migrated 70 services from a hand-managed VM fleet to Kubernetes on AWS over 14 months with no customer-facing downtime; the ops team's weekend work went to zero.
- Rebuilt CI on GitLab with cached, parallel pipelines; median build time fell from 34 to 9 minutes across 120 repositories.
- Wrote the Terraform modules for networking, databases and IAM used by every team; configuration drift incidents stopped after import of the last manual resource.
- Introduced centralised logging and tracing (Loki, Tempo, Grafana); time to find the failing service in an incident fell from about an hour to minutes.
- Automated server provisioning with Ansible for 9 client environments, replacing runbooks; environment setup went from three days to two hours.
- Ran backups, patching and monitoring for 300 Linux hosts with a 99.95% uptime record over two years.
Education
Shown in the Metro template
Questions
The one in the posting you are answering. The work overlaps; the headline can say 'DevOps / Platform Engineer' and the bullets should show reliability, delivery and cost, which all three roles are measured on.
Use the platform's own measures: deploy frequency, build time, incidents, uptime, cost, environments provisioned, teams onboarded. You can read every one of them from your tooling.
No. Group the tools you could be interviewed on today into four or five lines: platform, cloud and IaC, delivery and observability, languages. Thirty logos read as a list of things you know shallowly.
They pass filters and prove the basics; they do not replace a migration or a cost story. Keep CKA, AWS or GCP professional-level certificates in a dated section and spend the space on results.
More examples
All examples →Backend engineer with six years building payment and logistics services in Go and TypeScript. Owned services handling 40,000 requests per minute, cut a core API's p95 latency by 70%, and led the migration of a monolith into twelve services without a customer-facing incident. Comfortable on call, in design reviews and in front of a whiteboard with product.
Experience
- Led the split of the order monolith into twelve Go services behind an API gateway; deploy frequency went from weekly to 30+ releases a day with zero Sev-1 incidents during the migration.
- Rewrote the rate-calculation API with cached carrier tables and request coalescing; p95 latency dropped from 820 ms to 240 ms at 40,000 requests per minute.
- Designed the event schema and outbox pattern used by 9 teams; duplicate shipment events fell to zero after rollout.
- Ran the on-call rotation for the platform group and cut paged alerts by 60% by deleting noisy checks and adding SLO-based burn-rate alerts.
- Built the direct-debit mandate service (TypeScript, PostgreSQL) that processed £1.2bn in its first year; passed the external PCI DSS audit with no findings.
- Introduced contract tests between the mobile app and the ledger API; integration bugs found in production fell from 11 to 2 per quarter.
- Mentored three graduate engineers through their first production launches; all three were promoted within 18 months.
- Delivered React and Node features for eight client projects, from a council booking system to a retailer's loyalty app.
- Set up the agency's first CI pipeline (GitHub Actions, Docker), cutting release day from a full afternoon to 20 minutes.
Projects
An open-source job queue on PostgreSQL with exactly-once delivery; 2,100 GitHub stars and used in production by at least four companies.
Education
Skills
Certifications
Languages
Summary
Data analyst with five years in e-commerce and subscription businesses, turning warehouse data into decisions for product and marketing teams. Built the self-serve dashboards 140 people use weekly, found the pricing leak that recovered €1.4M a year, and runs the experiment readouts for twelve teams. Fluent in SQL and Python, and in the harder skill of saying what the data does not show.
Experience
- Found a discount-stacking leak in the checkout data and quantified it at €1.4M a year; the fix shipped within a month and recovered 92% of it.
- Built the Looker dashboards for product and growth (dbt models on BigQuery) used weekly by 140 people; ad-hoc data requests fell by 60%.
- Run experiment readouts for 12 teams and 90+ A/B tests a year; a shared power calculator and template cut tests stopped early on noise to near zero.
- Modelled customer lifetime value by acquisition channel; marketing moved 25% of paid budget to the two channels with the highest 12-month LTV.
- Built the churn-risk model (logistic regression in Python) that flags subscribers 30 days out; retention campaigns on the flagged group lifted saves by 18%.
- Rebuilt the weekly executive report from 40 manual spreadsheet steps to one dbt pipeline; preparation time went from a day to 20 minutes.
- Analysed the paywall meter with cohort analysis; raising the free-article limit from 3 to 5 increased trials by 22% without losing subscribers.
- Cleaned and analysed survey data for 20 client studies in R and SPSS; produced the charts and tables that went into every client report.
- Automated the monthly tracker report with R Markdown, saving the team roughly 30 hours a month.
Projects
A public dashboard tracking asking rents by area from 60,000 scraped listings; cited by two national newspapers.
Education
Skills
Certifications
Languages
Mechanical engineer with six years designing heat-transfer and material-handling products from concept to production. Designed a six-size heat exchanger family that cut unit cost 18%, took a compressor mount's field failure rate from 2.3% to 0.4% across 12,000 units, and cut prototype lead time from six weeks to nine days. EIT, Certified SolidWorks Professional, fluent in ANSYS and GD&T.
Experience
- Designed a six-size brazed-plate heat exchanger family in SolidWorks; part count fell from 41 to 27 and unit cost dropped 18% against the predecessor line.
- Led the DFMEA and tolerance stack-up for a new compressor mount; field failures fell from 2.3% to 0.4% across 12,000 units shipped in the first year.
- Ran CFD in ANSYS Fluent on a manifold redesign; pressure drop fell 22% and the test rig confirmed the model within 5%.
- Cut prototype lead time from six weeks to nine days by qualifying an in-house SLS printer and a local CNC vendor for first articles.
- Owned drawings and BOMs for a modular conveyor line of 300+ assemblies; engineering-change turnaround fell 40% after I moved the team onto SolidWorks PDM.
- Redesigned the drive-module gearbox mount; assembly time fell from 45 to 28 minutes, worth about $90,000 a year at 2,000 units.
- Supported seven customer installations on site and wrote the installation manual still shipped with every line.
- Designed welding fixtures for a sprayer boom; three of four were adopted on the line and cut setup time by about 15 minutes per unit.
- Ran FEA in ANSYS Mechanical on a hitch bracket and proposed a gusset that lifted the safety factor from 1.6 to 2.4 at no added weight.
Education
- Senior design: a low-cost solar water heater tested at 62% collector efficiency; team won the department's design award.