Coding Project Assessments That Mirror the Real Job
Realistic, AI-native coding projects that drop candidates into a real codebase with the AI tools they would use at work. Watch how they actually solve the problem, and know how they will perform on the job before you make an offer.
A Real Development Environment, Not a Coding Puzzle
Fully Functional VS Code, in the Browser
Candidates work in a fully functional VS Code environment in the browser: a real multi-file project where they can debug, install extensions, and run the code, exactly like their daily setup.
Embedded AI Tools
Candidates work with AI tools such as Claude Code or Codex exactly as they would on the job, and every AI interaction is captured for your team to review.
Automated Test Cases
Objective correctness is scored automatically, so every submission is measured against the same bar no matter who reviews it.
Beyond Technical Skills
Score the coding project alongside the rest of your assessment, so one scorecard covers code quality, problem solving, communication and adaptability.
See a Coding Project Assessment in Action
See How Candidates Really Work With AI Tools
Candidates complete realistic coding projects in a real development environment while using AI tools such as Claude Code or Codex, just as they would at work. See exactly how they use AI, how they validate its output, and how effectively they solve complex engineering challenges. All before you invest hours interviewing them.
Identify the difference between engineers who become 2× with AI and those who become 10×.
Objective Auto-Scoring, at Scale
Every project is evaluated using automated test cases for objective correctness, plus specialized AI agents that assess engineering skills beyond whether the code simply works. Measure what truly matters: how candidates prompt AI tools, when they rely on AI, how they validate AI-generated code, and whether they improve or blindly accept AI suggestions.
You get objective scores, detailed breakdowns, and interview insights, so your team knows exactly what to explore in the next stage.
Set Your Own AI Usage Policy
Candidates know exactly what AI assistance is allowed for every project. Configure your own AI usage policy, from unrestricted AI collaboration to controlled usage, so every candidate is evaluated fairly and consistently.
What Is a Coding Project Assessment?
A coding project assessment is a realistic take-home engineering assignment. Instead of solving an isolated puzzle, the candidate works inside an existing multi-file project in a fully functional VS Code environment in the browser, where they can debug, install extensions, and use the AI tools your policy allows. They produce work that resembles a normal day on your team: automated test cases score objective correctness, and specialized AI agents assess the engineering skills behind the result, so you see how someone will perform on the job rather than how well they memorised an algorithm.
100% Customizable. Built Around Your Engineering Team.
Your Tech Stack
Set the languages and runtimes your team actually ships in, so the project matches the codebase a new hire would join.
Your Frameworks
Build the project on the frameworks and libraries your engineers work with every day, rather than a generic exercise.
Your AI Tools
Choose which AI assistants candidates can work with, such as Claude Code, Codex or your own preferred tooling.
Project Complexity
Scale the scope of the project to the role, from a focused fix for a junior hire to an architectural change for a senior one.
Scoring Criteria
Decide what the score rewards, from correctness and code quality to how well the candidate collaborates with AI.
AI Usage Policy
Set how much AI assistance is allowed per project, from unrestricted collaboration to controlled usage.
Find Out Who Can Actually Do the Job
Frequently Asked Questions
What is a coding project assessment?
A coding project assessment is a realistic take-home engineering assignment. Instead of solving an isolated puzzle, the candidate works inside an existing multi-file project in a fully functional VS Code environment in the browser, where they can debug, install extensions, and use the AI tools your policy allows, producing work that resembles a normal day on your team.
Can candidates use AI tools during a coding project?
Yes, and that is the point. Candidates can work with AI tools such as Claude Code or Codex, and you decide how much assistance is allowed for each project, from unrestricted AI collaboration to controlled usage. Every candidate sees the policy up front, so results stay comparable.
How are coding projects scored?
In two layers. Automated test cases score objective correctness, and specialized AI agents assess the engineering skills behind the result: how the candidate prompted AI tools, when they relied on AI, how they validated AI-generated code, and whether they improved or blindly accepted AI suggestions. You get a score, a detailed breakdown, and interview insights for the next stage.
How is a coding project different from a live coding interview?
A coding project is asynchronous: the candidate works alone, in their own time, on a larger piece of work. A live coding interview happens in real time with an interviewer in the same environment. Many teams use both, a coding project to screen for real delivery ability and a live coding interview to explore the reasoning behind it.
Which languages, frameworks and tech stacks can I use?
Coding projects are fully customizable. You choose the tech stack, frameworks, AI tools, project complexity, scoring criteria and AI usage policy, so the project matches the work your engineers actually do.
Do coding projects work for junior and senior engineers?
Yes. Because you set the project complexity and the scoring criteria, the same format works for a junior role and for a senior hire. The signal changes with the project you choose.
Can I combine a coding project with the rest of my assessment?
Yes. Coding project results are scored in the same platform as the rest of your assessment, so technical output, soft skills and AI collaboration land on one combined scorecard and candidate ranking.
Can I integrate coding project assessments with my ATS?
Yes. Canditech integrates with 40+ ATS platforms including Lever, Greenhouse, Workable, SmartRecruiters, Workday and SAP SuccessFactors, so invites and results flow through your existing hiring workflow.
