Codility

Codility is a technical assessment platform used in engineering hiring. Its function is to give a hiring team evidence about a candidate's ability to write working software before committing to expensive interview time, and it does this with automated tests that are scored objectively rather than by impression. The company has been in this market since around 2009, which makes it one of the longer-established players, and its scoring approach, based on both correctness and computational performance, reflects an origin in algorithmic evaluation rather than in the more recent fashion for realistic work simulation.

The core product is the coding test. A candidate receives a link, opens a browser-based editor and solves one or more problems within a time limit. Solutions are executed against a set of test cases, and scoring is automatic. What distinguishes Codility's historical approach is that a solution is evaluated on two axes: whether it produces correct output for all test cases, and whether it does so within performance constraints. A solution that is correct but quadratic when a linear one is expected scores lower. This catches a specific and important engineering quality, the ability to reason about how code behaves at scale, which a correctness-only test misses and which a human reviewer often fails to assess reliably in a take-home submission.

The test library is maintained by the vendor and covers algorithms, data structures, language-specific problems, SQL, and role-oriented assessments for frontend, backend, mobile, data, quality assurance and other specialisms. Custom questions can be added, and some teams use their own codebase as the basis for an assessment, which improves signal at the cost of setup effort and the risk of leaking internal code to candidates.

Real-life tasks and projects represent the platform's response to a well-founded criticism of algorithmic testing. The objection is that inverting a binary tree on a whiteboard has little relationship to the work of building and maintaining software, that it disadvantages experienced engineers who have not practiced puzzles recently, and that it correlates with familiarity with a particular style of interview more than with job performance. Codility addresses this with assessment formats closer to actual work: debugging an existing codebase, extending a small application, refactoring poor code, writing tests, fixing a failing build, or completing a take-home project with a longer time window and a submission reviewed by the hiring team. These formats produce better signal for senior roles and reduce adverse impact, and most mature hiring teams now blend a short algorithmic screen with a realistic task.

Proctoring and integrity features have become central to this category, and their importance increased sharply as generative AI made it trivial for a candidate to paste a problem into a model and submit the answer. Codility's approach includes monitoring signals during an assessment: tab switching and window focus loss, copy and paste behaviour, unusual typing patterns suggesting pasted code, timing anomalies, and in some configurations webcam or screen capture. Post-submission analysis can compare the solution's development pattern against what a genuinely authored solution looks like, and similarity detection flags code matching known sources. Some products in this space now attempt direct detection of AI-generated solutions, which is an arms race rather than a solved problem; a buyer should ask what is actually measured, what false positive rates look like, and how a flagged candidate is handled, because accusing a good candidate of cheating on weak evidence is a costly error.

Live interviews are supported with a shared coding environment, so an interviewer and a candidate can work in the same editor with the session recorded for later review by other interviewers. This addresses a real problem in hiring loops where each interviewer re-asks similar questions and no one has visibility into what was already covered. Structured scorecards help standardise evaluation across interviewers.

Integration with applicant tracking systems connects assessment results to the hiring pipeline, so scores appear against a candidate record, invitations are triggered by stage changes, and reporting can cover the funnel end to end. This is what makes the platform usable inside an existing recruiting process rather than as a side system. Reporting and benchmarking allow comparison of candidates against each other and, where the vendor's data permits, against a broader population, and they support analysis of whether assessment scores actually predict on-the-job performance, which is the question a thoughtful hiring manager should be asking.

TestGorilla acquired Codility, combining a broad pre-employment screening platform covering cognitive, personality, situational judgment and role-specific tests with Codility's deep technical assessment capability. For a buyer this changes the commercial conversation: technical screening can now be purchased as part of a wider assessment suite rather than standalone, which can simplify procurement and reporting, and it also means the roadmap is set by a larger organisation whose centre of gravity is broader than engineering hiring. Current product packaging and terms should be confirmed with the vendor.

Against competitors, HackerRank is the largest and best-known rival, with a bigger brand, a very large question library, a free tier for individual developers that creates a network effect, skills certifications candidates can present, strong interview tooling and broad language coverage. HackerRank generally wins on ecosystem and recognition; Codility has historically won on scoring rigour for algorithmic performance and on a more focused product. TestGorilla's own platform competes on breadth of non-technical assessment at accessible pricing. CodeSignal offers a certified general assessment score intended to be reusable across employers, plus realistic simulations, which is a different structural idea. CodeSubmit, TestDome and Qualified focus on realistic work-sample tests and take-home projects. DevSkiller and Harver serve adjacent segments. HireVue specialises in video interviewing with AI-assisted analysis. On the open-source and self-run side, many engineering teams build their own take-home assignments and pair-programming interviews, which produce better signal per dollar for small hiring volumes at the cost of engineering time and inconsistency. Codility's position is a mature, rigorous, technically credible assessment platform now inside a broader screening group.

Pricing is quote-based and subscription-driven, typically billed annually and structured around recruiter seats, hiring manager access and annual assessment volume. Tiers differentiate interview capabilities, proctoring depth, ATS integration, reporting, custom questions, support level and administrative controls. A free trial allows evaluation. Because pricing is not published, comparing cost against HackerRank or TestGorilla requires sales conversations with each, and buyers should negotiate on volume commitments and multi-year terms. For a company hiring a handful of engineers a year, the cost of a platform may exceed the cost of running a good in-house take-home process; for a company hiring continuously, the automation pays for itself many times over.

The limitations should be stated honestly. Algorithmic testing, however rigorously scored, predicts job performance weakly for many roles, and over-reliance on it filters out capable experienced engineers who do not practice puzzles. Assessment fatigue is real: strong candidates with options will abandon a process that begins with a ninety-minute test, so assessment length and placement in the funnel matter as much as assessment quality. Proctoring creates an adversarial tone that can damage employer brand, and invasive monitoring raises privacy questions under European and other data protection regimes that require a lawful basis and transparency. AI detection is not reliable enough to be the sole basis for rejecting a candidate. Benchmarks derived from vendor populations may not resemble your candidate pool. And a platform is a tool, not a hiring strategy: a team with unclear role requirements and unstructured interviews will make bad decisions faster with better data.

Codility fits companies hiring engineers continuously at volume, organisations that need objective, defensible screening evidence with integrity controls, teams building structured hiring processes with scorecards and calibrated interviews, employers screening for roles where algorithmic and performance reasoning genuinely matters, and groups that want technical assessment inside a broader pre-employment screening purchase. It does not fit companies hiring one or two engineers a year, roles where a work-sample portfolio or a paid contract task is more appropriate, employers unwilling to invest in structured interviewing around the assessment, or organisations with strong privacy constraints on candidate monitoring.

In the closing comparison, choose Codility when scoring rigour, proctoring and integration into a structured technical hiring process are the requirements, particularly within a TestGorilla-based screening strategy. Choose HackerRank when brand recognition, library breadth, certifications and a developer-facing ecosystem matter more. Choose CodeSignal when you want a reusable certified score or realistic role simulations. Choose CodeSubmit or TestDome when work-sample realism is the priority and you hire at lower volume. Choose an in-house take-home and pair-programming process when you hire rarely and can invest engineering time. The most defensible approach for a serious engineering organisation is a short screen early, a realistic work sample later, structured interviews throughout, and a regular review of whether assessment scores actually correlate with the performance of the people you hired.