Codebase navigation, architectural understanding and making sense of ambiguous requirements are invaluable skills in any programmer's toolbelt.
With thousands of datapoints to analyze, from quality of clarifying questions, to navigation efficiency, we generate an in-depth, explainable analysis of every candidate
Save hundreds of hours spent by engineering teams on multiple interview rounds with our AI interview companion. Utilize reports, proctoring and replay to make a decision in a tens of minutes instead of hours.
With multiple integrations available and no external dependencies, StackLitmus fits right into your hiring pipeline with minimal effort
Problems of varying difficulties are carefully designed and presented as expected business outcomes rather than explicit instructions. This allows a more accurate gauge of problem solving ability and encourages judgement, communication and iteration, which are natural phases in the software development lifecycle.
All test scenarios are wrapped in a codebase designed to mimic real world abstractions, complexity and varying levels of technical debt. The candidate is expected to navigate, understand and eventually find the place where a change needs to be made to meet the required outcome of the test.
Preset requirements are revealed by the AI assistant to the candidate gradually, with minor changes and major changes being requested to simulate the changing needs of a project along with the candidate's experience and agility. The candidate is allowed to question, clarify and to finally submit changes.
Based on the candidate's performance in these assessments, a heuristic-based explainable analysis is generated that highlights areas of interest in programming proficiency. In order to achieve this, StackLitmus calculates code quality, algorithmic efficiency, communication effectiveness and correctness of results
Our assessments are designed to test a candidate's interaction with the codebase and with the AI assistant for clarifying questions, and not just the final result. As a result, Stacklitmus monitors and saves all the actions taken to provide a replayable report of behavioural anomalies that are detected by our system. We ensure that the candidate navigates the codebase and finds solutions with a natural path and at a natural pace.
Customize and view any of these sandboxes to get a feel for what our tests look like