New Video Scoring Feature Added to AI Automated Scoring System 'AI.R-Scorer,' Available from April 28
Level Enter Co., Ltd. has added a video scoring function to its AI automated feedback system 'AI.R-Scorer'. By analyzing videos of presentations or interviews, the AI provides objective evaluations of non-verbal elements like speaking style and gestures, supporting qualitative assessment in education and recruitment.
📋 Article Processing Timeline
- 📰 Published: April 28, 2026 at 19:00
- 🔍 Collected: April 28, 2026 at 10:32
- 🤖 AI Analyzed: April 28, 2026 at 14:14 (3h 42m after Collected)
Level Enter Co., Ltd. (Headquarters: 5-32-12 Shiba, Minato-ku, Tokyo; CEO: Dai Yamamoto), a company engaged in service planning, development, and programming education, will add a new feature to score videos to its AI-powered automated scoring and feedback system 'AI.R-Scorer' (originally released in 2024 for essays and statements of purpose) and will begin providing it on April 28, 2026. This function was developed to provide objective evaluations for videos such as presentations of inquiry-based learning results and interview practice. The following new features related to video scoring have been added:
- Automatically compress multiple selected videos and upload them at the optimal size (up to 5 videos can be uploaded at once).
- Score slide design and other elements based on the video.
- Score speaking style, gestures, and other factors based on the video.
As part of function verification, videos of actual result presentations were scored, and features were implemented based on feedback from active teachers. Scoring these high-capacity videos was made possible by the 'background parallel scoring function' added in the February 2026 update, which allows other tasks to be performed during processing. It can be used for evaluating inquiry learning results, assessing interview practice, and evaluating teachers' classroom performance. By evaluating voice, facial expressions, and movements directly from the video itself rather than just transcripts, the system aims to achieve 'fair and consistent qualitative evaluation' beyond quantitative analysis of slide information.
### Development Background
In inquiry-based learning, which is gaining importance in education, the final evaluation often involves a presentation. Points to be evaluated are broad, including perspectives, methods, organization, slide content, and demeanor. Securing time for multiple evaluators to attend in person was a challenge. Educational institutions requested improvements regarding operational load and scoring inconsistencies, leading to this development. Beyond inquiry learning, it can be used for:
- Evaluating interview or presentation practice to improve response accuracy.
- Evaluating teachers' classroom scenes to improve teaching techniques.
Furthermore, by allowing individual settings for detailed evaluation indicators, the system contributes to the realization of 'AI-based qualitative evaluation' requested by companies.
### Features of the New AI.R-Scorer Function
The video scoring function has been added while maintaining existing workflows. Based on needs from users of Ver. 2.0 (2024) and Ver. 3.0 (February 2026), the specifications focus on solving practical problems in evaluation work. It supports consistency and fairness in qualitative evaluation by directly assessing the video itself.
**Key Features:**
- Evaluation covers speech demeanor (voice tone, gestures, etc.) as well as slides.
- Automatic compression before upload.
- Automatic splitting of videos by subject when batch selected.
- Customizable evaluation indicators and criteria.
**Actual Case:**
A presentation on the 'Cam-Cam Project,' which aims to eliminate barriers for people with dysphagia, was evaluated. The presentation covered three years of background, results, and future prospects. It was comprehensively evaluated for slide structure, logic, and expressive power, receiving a score of 89 points.
- Automatically compress multiple selected videos and upload them at the optimal size (up to 5 videos can be uploaded at once).
- Score slide design and other elements based on the video.
- Score speaking style, gestures, and other factors based on the video.
As part of function verification, videos of actual result presentations were scored, and features were implemented based on feedback from active teachers. Scoring these high-capacity videos was made possible by the 'background parallel scoring function' added in the February 2026 update, which allows other tasks to be performed during processing. It can be used for evaluating inquiry learning results, assessing interview practice, and evaluating teachers' classroom performance. By evaluating voice, facial expressions, and movements directly from the video itself rather than just transcripts, the system aims to achieve 'fair and consistent qualitative evaluation' beyond quantitative analysis of slide information.
### Development Background
In inquiry-based learning, which is gaining importance in education, the final evaluation often involves a presentation. Points to be evaluated are broad, including perspectives, methods, organization, slide content, and demeanor. Securing time for multiple evaluators to attend in person was a challenge. Educational institutions requested improvements regarding operational load and scoring inconsistencies, leading to this development. Beyond inquiry learning, it can be used for:
- Evaluating interview or presentation practice to improve response accuracy.
- Evaluating teachers' classroom scenes to improve teaching techniques.
Furthermore, by allowing individual settings for detailed evaluation indicators, the system contributes to the realization of 'AI-based qualitative evaluation' requested by companies.
### Features of the New AI.R-Scorer Function
The video scoring function has been added while maintaining existing workflows. Based on needs from users of Ver. 2.0 (2024) and Ver. 3.0 (February 2026), the specifications focus on solving practical problems in evaluation work. It supports consistency and fairness in qualitative evaluation by directly assessing the video itself.
**Key Features:**
- Evaluation covers speech demeanor (voice tone, gestures, etc.) as well as slides.
- Automatic compression before upload.
- Automatic splitting of videos by subject when batch selected.
- Customizable evaluation indicators and criteria.
**Actual Case:**
A presentation on the 'Cam-Cam Project,' which aims to eliminate barriers for people with dysphagia, was evaluated. The presentation covered three years of background, results, and future prospects. It was comprehensively evaluated for slide structure, logic, and expressive power, receiving a score of 89 points.