Getting into Gauntlet AI isn't about where you worked or what school you went to. It's about output, speed, and mental sharpness. The application process is designed to filter for people who are not just smart, but can execute. You don’t need to know everything about AI before applying. But you do need strong software development fundamentals and a strong cognitive aptitude.
Here's what to expect and how to prepare.
The Application Process – 4 Steps
1. Upload Resume & Personal Info
The first step is simple but essential.
You’ll be asked to:
Upload your resume
Enter basic personal and contact info
Share optional URLs (LinkedIn, portfolio, etc.)
Once your resume is uploaded, our system uses AI to auto-fill your profile — you’ll just need to confirm it’s accurate before proceeding. From here, our team quickly evaluates your background to determine if you’re ready to move to the next step.
🕐 Expect a response via email within the hour.
If you meet our minimum experience threshold (typically 3+ years of hands-on development or equivalent self-taught output), you’ll be invited to take the CCAT.
2. Take the CCAT (Cognitive Aptitude Test)
The Criteria Cognitive Aptitude Test (CCAT) is a 50-question, 15-minute exam that evaluates raw intelligence: problem-solving, pattern recognition, logic, and verbal reasoning.
You get two chances to take the test – your highest score counts.
Most people don’t finish all 50. That’s by design. The average score is around 24. Gauntlet applicants must score 40 or higher, which is the 98th percentile. We set the bar that high because we want engineers with elite cognitive horsepower — the kind of thinking it takes to build and adapt quickly in an AI-first world.
Topics Covered:
Critical thinking & logic
Math & numerical patterns
Verbal reasoning
Abstract/spatial analysis
How to prep:
We recommend timed drills, AI-generated CCAT-style questions, and focused practice to build speed. We wrote a blog about CCAT preparation here.
3. Technical Coding Exam
If you pass the CCAT, you’ll receive an invite to our coding exam – a short technical screen designed to measure fluency, fundamentals, and debugging ability.
This is a LeetCode-style challenge with:
3 questions, increasing in difficulty
5 minutes per question once opened
Up to 3 code submissions per problem
Syntax highlighting & full code editor
Choose from 10+ languages (Python, JavaScript, Java, Go, Rust, C++, etc.)
This isn’t about academic trickery. It’s about seeing how fast you can move, how cleanly you code, and how well you solve real-world problems.
⏱️ Prep tip: Practice medium/hard LeetCode problems and simulate timed rounds.
4. Await Next Steps (Final Review)
Once you’ve pass both tests, our team will review your application. If you pass the thresholds for both CCAT and the technical exam, we’ll follow up with the next phase. If you made it this far, you are in a great spot.
We typically notify candidates within 48–72 hours of completion.
Other things to know:
- You can only apply once per cohort. If you don’t pass, you can apply to the next cohort.
- While the current tests are not proctored, you will take a proctored version later in the program. Gauntlet compares results to confirm consistency.
- Cheating is flagged automatically, even if it’s not announced. Don’t bother trying.
What Comes Next: Remote Trial Work
If you pass the initial screening, you’re invited to complete remote trial work. This is the real test.
- 80 to 100 hours over 3 weeks
- Projects with deadlines, technical requirements, and stretch goals
- Everything is scored and compared across applicants
What to Know Before Applying
- You don’t need to be an AI expert
- You do need to be good at writing software and thinking clearly
- Practice the CCAT and LeetCode now
- Be ready to grind for 100 hours if you pass
Final Notes
This application is a filter. It’s supposed to be hard. If you don’t pass, study and reapply for the next cohort. Gauntlet is building something serious and we’re not lowering the bar. We hope to see you Inside the Gauntlet.
Click here to apply to Gauntlet AI.