About
Engineering ability, made visible.
A place for engineers to develop and demonstrate their skills, and for teams to recognize the ability behind the work.
01 / Mission
Our mission.
Help people build their skills and prove what they’re capable of.
Today, we serve that purpose through engineering practice, asynchronous assessments, and live interviews. Developers get room to build skills, test ideas, and explain their decisions. Teams get work, results, and reasoning they can review together. As tools change, the purpose stays the same: help people grow and make their abilities visible.
02 / Vision
Our vision.
A world where everyone has a fair chance to find work and build a better future.
We believe everyone deserves a fair chance to get a job and grow. That starts with room to learn and a way to show what they can do, wherever they started or learned their skills. Practical tasks, clear expectations, and reviewable work help teams recognize ability. People remain accountable for hiring decisions.
03 / Values
Our values.
Practical skills
Coding, debugging, AI code review, and bug finding give engineers concrete work to practice, demonstrate, and discuss.
AI with accountability
Using AI still calls for engineering judgment. We make room to question output, test assumptions, and explain trade-offs.
A fair chance to show your ability
We want people to have a fair chance at a job, wherever they started or learned their skills. Clear tasks and reviewable work help candidates demonstrate what they can contribute.
Evidence people can inspect
Tests, code history, reasoning, and available integrity signals give reviewers context. People remain accountable for the hiring decision.
04 / Results
How we keep results useful.
Consistent conditions
Candidates see the same instructions, time limits, task types, and allowed-tool rules for a given technical screen.
Visible evidence
Session replay, code history, execution results, and integrity signals make the result inspectable instead of a black box.
AI-era judgment
We test whether candidates can use AI critically: direct it, challenge it, fix it, and verify the result.
Human review
Scores and benchmarks support the hiring team. Final decisions belong to people who understand the role and context.
05 / Product
What we offer.
Technical Screens
Run role-specific technical screens across coding, debugging, AI code review, and bug finding work.
AI code review
The screen that measures whether candidates can catch confident-but-wrong AI code and hold their ground when it pushes back.
Verification Mode
Hand candidates AI-generated code and see who can actually find the bugs versus who blindly trusts the output.
Code Replay
Watch how a candidate approached each problem, edit by edit, with test runs and final code preserved.
Discernment Evidence
Reviewable evidence showing whether a candidate caught real defects, avoided false alarms, and expressed uncertainty where captured.
Integrity Events, Built In
Focus changes, paste events, run history, and replay give reviewers concrete context without face, voice, or emotion scoring.
22 Languages
Assess in Python, JavaScript, TypeScript, Java, C, C++, C#, Go, Rust, Ruby, PHP, Kotlin, Swift, Scala, Dart, Haskell, Elixir, Clojure, Groovy, Lua, Perl, or SQL with full code execution. Each task offers only the languages its test harness runs exactly.
Custom Screens
Mix coding, verification, prompting, and live validation tasks into screens tailored to each role.
Are you a developer? You can sharpen the same skills in Practice.
06 / Origin
Why we built this.
AI can produce code quickly. Engineers remain responsible for understanding the problem, choosing an approach, checking the result, and explaining the trade-offs.
We built DevEval around that responsibility. Developers can practice the work, reflect on feedback, and try again. Hiring teams can set clear conditions, inspect what candidates did, and use live follow-up to explore what the record leaves uncertain.
The purpose is shared: help engineers show their ability and help teams recognize it.
07 / Questions
Questions? We have answers.
What is DevEval?
DevEval combines public engineering practice with technical assessments and live interviews. Developers work on coding, AI code review, and bug finding; hiring teams review task results, candidate reasoning, and available session evidence.
How is this different from a traditional coding test?
Traditional tests over-index on producing a passing solution. DevEval keeps executable fundamentals, then adds judgment in the AI era: whether candidates can catch confident-but-wrong AI code, verify output, and direct an assistant well. Reviewers inspect that evidence alongside the rest of the hiring loop.
How do candidates join a screen?
You create a technical screen and send invite links by email. Candidates open the link in the browser, no account required. Results, including code replay and integrity events, appear in your dashboard.
Can developers use DevEval to practice?
Yes. Developers can sharpen the same skills in Practice, working through interview-style problems across 22 languages. It is a secondary part of the product; our focus is helping companies hire.
How much does it cost?
New workspaces require a paid plan to invite candidates. Paid plans start at $59/month billed annually, or $79 billed monthly, and every plan includes unlimited team members. Business and Enterprise add more sessions, integrations, and dedicated support. See pricing or talk to sales.
08 / Product feedback
Help us make DevEval sharper.
Choose the signal that best matches what you found. Each path asks only for the context that helps our team act on it.
DevEval447 Sutter St, Ste 506-1440
San Francisco, CA 94108
Talk to us
Bring one hiring need.
A 15-minute walkthrough with your open role.