Frequently Asked Questions
Is GradeSense allowed for board exams / competitive exams?
Yes, GradeSense is designed to assist evaluators, not replace them. For high-stakes exams, we recommend our "Human-in-the-Loop" workflow where AI provides a first pass, and human experts verify 100% of the scripts.
Do teachers lose control when using AI grading?
Absolutely not. Teachers have full control to override any AI suggestion. The AI acts as a smart assistant that handles the repetitive work, allowing teachers to focus on complex judgments.
How accurate is GradeSense compared to human graders?
In our benchmarks, GradeSense shows ~90% agreement with experienced human graders, which is comparable to (or sometimes better than) the agreement rate between two different human evaluators.
Can we start with a small pilot on one subject?
Yes! We encourage a pilot approach. You can start with a single subject or exam batch to validate the workflow before rolling it out institute-wide.
Where is our data stored?
All data is hosted securely in India (Mumbai AWS/GCP regions), ensuring compliance with local data residency norms.
Do you train your models on our answer sheets?
No. We do not use your students' answer sheets to train our foundation models or share them with other institutions. Your data improves your instance of GradeSense only.
What if the AI makes a mistake?
The AI highlights "low confidence" scores for manual review. Additionally, teachers can easily spot-check and override any grade in the "Review" interface. The system learns from these corrections.
How long does it take to get started?
For a pilot, we can get you up and running in less than 48 hours.
Does GradeSense support multiple languages?
Yes, we support English, Hindi, and select other regional languages for both handwriting recognition and semantic grading.
Can we integrate with our existing LMS?
Yes, we offer custom integrations for major LMS and SIS platforms on our Enterprise plan.