Three assessment formats. One evaluates how candidates reason alongside AI and logs every prompt they use. All auto-graded, all integrated into your hiring pipeline.
96%
avg correctness score
3
assessment formats
15+
languages supported
5 min
to build and publish
The blank editor test was invented before Stack Overflow existed.
Engineers who can reason through real problems with AI ship faster than those who can only recall syntax from memory. Your assessments should measure that, not penalize it.
87%
of engineers use AI tools daily
5 min
to build a full assessment
4.8/5
rating on G2
One platform. Three ways to evaluate.
Pick the format that fits the role. Combine multiple types in a single assessment flow. All auto-graded and integrated into your pipeline.
1def find_minimum_path(grid):2 rows, cols = len(grid), len(grid[0])3 dp = [[float('inf')] * cols for _ in range(rows)]4 dp[0][0] = grid[0][0]5 for i in range(1, rows):6 dp[i][0] = dp[i-1][0] + grid[i][0]7 for j in range(1, cols):8 dp[0][j] = dp[0][j-1] + grid[0][j]9 for i in range(1, rows):10 for j in range(1, cols):11 dp[i][j] = min(dp[i-1][j], dp[i][j-1]) + grid[i][j]12|
What's the difference between bottom-up and top-down DP?
How do I handle an edge case for a 1x1 grid?
Write the complete solution for me
What time complexity does my current loop give?
4 prompts logged
Give candidates a built-in AI assistant during the coding test. Every prompt they send is logged. You see how they think, not just whether they got the answer right.
When to use: Senior engineering roles, full-stack positions, any role where real-world problem solving matters more than algorithm recall.
- Built-in AI assistant, not external tools
- Full prompt history logged for review
- Reasoning score alongside code score
- 15+ programming languages supported
You don't just see the answer. You see how they got there.
Vibe AI logs every prompt a candidate sends during the assessment. Review their full reasoning process, not just the code they submitted. The difference between using AI as a crutch and using it as a thinking tool is visible in the prompt log.
Full prompt history
Every question the candidate sent to Vibe AI, timestamped and categorized by signal type.
Reasoning over output
A candidate who asks the right questions is more valuable than one who got lucky with the right answer.
Automatic red flags
Prompts showing overreliance on AI are surfaced immediately so you can review them in context.
Assessment Results
Priya Sharma — Senior Backend Engineer
87
overall
AI Prompt Log
4 promptsHow do I handle concurrent requests in Node.js?
Good signalWrite the complete solution for me
OverrelianceWhat's the time complexity of my current approach?
Self-reviewHow do I handle the empty array edge case?
DebuggingNot just a score. The full evaluation record.
Every completed assessment generates two deliverables automatically: a screen recording of the coding session and a written AI evaluation report. Your reviewers get context, not just a number.
Screen recording of the full session
Captures every keystroke, test run, and Vibe AI interaction. Reviewers can watch the candidate's problem-solving process, not just the final output.
AI evaluation report with narrative
A written assessment covering code quality, reasoning patterns, and how the candidate used Vibe AI. Generated the moment they submit.
Ready to download and share
Export the report as a PDF or share a link. The recording and report travel together so stakeholders always have the full picture.
Recording and report generation is automatic. No setup required, no waiting — the moment a candidate submits, both are ready in your dashboard.
Assessment Report
Arjun Mehta
Backend Engineer Intern · Vibe AI assessment
87
overall
AI Evaluation
Arjun solved the minimum path problem correctly with O(m×n) time complexity and handled both edge cases without prompting. His Vibe AI prompt log shows 4 focused questions — all conceptual or debugging in nature, no requests for complete solutions. Code is clean, variable naming is clear, and the approach was explained coherently in comments. No performance or overreliance concerns identified.
Attachments
Screen Recording
42:18 · Minimum Path Sum
Evaluation Report PDF
3 pages · Generated by JIA
Everything you need to know.
What is Vibe AI and how is it different from allowing ChatGPT?
Vibe AI is a built-in AI assistant within JIA, not permission to use external tools. Candidates get a controlled, observable assistant for the duration of the assessment, and every prompt they send is logged for your review. This is fundamentally different from detecting or allowing external tools — it creates a structured environment where AI use is a first-class signal, not a variable to control out.
Can candidates cheat using external AI tools?
JIA does not block browser access, but Vibe AI assessments are designed for this reality. You are evaluating how candidates use AI, not whether they avoid it. Candidates who outsource their entire solution produce prompt logs and submissions that reveal this clearly. The signal is still useful.
What programming languages are supported?
JIA supports 15+ languages including Python, JavaScript, TypeScript, Java, Go, C++, Rust, Ruby, PHP, Kotlin, Swift, Scala, C#, R, and SQL. New languages are added based on customer demand.
How long does it take to create an assessment?
Most assessments are live in under 5 minutes. Choose a format, add your questions or select from the pre-built library, set time limits, and publish. Vibe AI assessments take the same amount of time to configure as a standard code test.
Can I use JIA assessments for non-technical roles?
Yes. MCQ assessments work for any knowledge domain, and Business Aptitude assessments are built specifically for product, operations, and leadership roles. You can combine formats in a single assessment flow for any role.
How are Vibe AI assessments scored?
Candidates receive a composite score across three dimensions: code correctness (test cases pass rate), code quality (structure and efficiency), and reasoning quality based on prompt log analysis. Each dimension is visible separately in the results view so you can weigh them based on your role requirements.
Can I add custom questions or use my own question bank?
Yes. You can write your own questions for any format, upload a question bank via CSV, or mix your custom questions with JIA's pre-built library. All questions are versioned and reusable across assessments.
Does JIA integrate with existing ATS systems?
JIA integrates with major ATS platforms. Assessment results sync directly into candidate profiles in your ATS. API access is available on Team and Enterprise plans for custom integrations.
Get started
See how Vibe AI changes the way you evaluate engineers.
Create your first Vibe AI assessment in under 5 minutes. Start with a $1.00 trial, no commitment required.