Canditech’s AI auto-scoring capabilities help you evaluate candidate responses faster, more consistently, and with greater objectivity. AI auto-scoring can be used for open-text questions, video questions, and coding questions, helping reduce manual review time while ensuring that every candidate is assessed according to the same clear standards.
In this article, we’ll walk through:
Note: AI Auto-Scoring is available on Enterprise plans. If you're interested in enabling it for your account, please contact our team.
Which Question Types Support AI Auto-Scoring?
AI auto-scoring is available for open-text, video, and coding questions.
For open-text and video questions, AI scoring evaluates the candidate's response based on predefined or custom criteria.
For coding questions, AI scoring can evaluate dimensions beyond functional correctness, including code quality, readability, time and space complexity, adherence to best practices, and custom requirements defined by the hiring team.
Note: For video responses, Canditech automatically converts the candidate’s speech into text and evaluates only the written transcription. Factors such as accent, tone of voice, appearance, background, or recording environment are never taken into account, allowing you to focus purely on what the candidate says rather than how they say it.
For more information on the question types and scoring methods available in Canditech, click here.
What AI Auto-Scoring Methods Are Available in Canditech?
Canditech currently offers two auto-scoring approaches, so you can choose the level of structure and customization that fits the role.
For more information on reviewing candidates' AI Auto-Scoring Results, click here.
1. Pre-Built AI Agents
Canditech provides several pre-built AI agents that are ready to use and require no configuration. Depending on the question type, different pre-built agents are available to help evaluate candidate responses.
Open-Text and Video Questions:
Language Agent - evaluates grammar, spelling, and overall language clarity in any language.
Objection Handling Agent - evaluates how candidates respond to objections in a sales context. This agent focuses on how well the candidate acknowledges objections, responds professionally, and moves the conversation forward. It is particularly relevant for sales, business development, and account management roles.
Conflict Resolution Agent - evaluates how candidates handle frustrated or dissatisfied customers. This agent focuses on empathy, communication, and problem ownership, making it especially suitable for customer support and customer success roles.
Coding Questions
Time Complexity Agent - evaluates the efficiency of a candidate's solution and whether the chosen algorithm demonstrates an appropriate time complexity for the problem.
Space Complexity Agent - evaluates memory usage and whether the candidate's solution uses resources efficiently.
For more information on reviewing candidates' AI Auto-Scoring Results, click here.
2. Custom AI Auto-Scoring (Custom Agents)
Canditech's Custom Agents are the most flexible and powerful AI auto-scoring option.
This option allows hiring teams to decide exactly what they want to evaluate in a candidate's response or coding submission by defining their own scoring rules. Instead of relying on generic criteria, you can translate your expectations, priorities, and evaluation standards directly into measurable checks.
Custom agents are especially useful when you want to align evaluation with your company culture and preferred ways of thinking or working, ensuring that candidate responses reflect your organization’s values. They are also ideal for addressing unique business challenges or technical requirements, allowing you to assess how candidates would approach scenarios that are directly relevant to your company's day-to-day operations.
For coding questions, custom AI agents can be used to evaluate specific rules defined by your engineering team, such as whether the solution uses recursion, avoids built-in sort functions, avoids duplicated logic, or uses meaningful variable and function names.
For an in-depth explanation on building your own Custom AI Auto-Scoring Agents, click here.
For more information on reviewing candidates' AI Auto-Scoring Results, click here.