Background: Challenges Unique to Educational Settings
While various generative AI models are being developed, university career centers face two major challenges:
1. **Security Risks:** Concerns over sending sensitive student personal information to external clouds. 2. **Communication Quality Barrier:** General-purpose AI often fails to reproduce the complex context of new graduate recruitment and the empathetic tone required of advisors.
Technical Features and Evaluation: Balancing Local Operation with Practical Quality
This model was trained exclusively on high-quality dialogues selected from over 10,000 job consultation records. It achieves structured Entry Sheet (ES) editing and 'empathetic dialogue' that aligns with actual practice. In comparative evaluations with major LLMs, while it may defer to large-scale cloud AI in terms of total knowledge, it has been confirmed to perform at an equivalent level in terms of 'ES editing quality' and 'dialogue tone.' It is a practical model ready for immediate use in local environments that protect confidential information.
Release on Hugging Face to Aim for Co-creation among Academia, Industry, and Government
To raise the standard of Japan's job-hunting support infrastructure and accelerate 'co-creation' among academia, industry, and government, we are releasing this model on Hugging Face.
* **Repository Name:** careerbot/shukatsu-gemma4 * **Ease of Implementation:** We adopted 'Gated Access,' which allows immediate access simply by agreeing to the terms of use. By utilizing tools such as Ollama and LM Studio, it is possible to verify and implement high-precision job-hunting support AI in a local environment.
*Note: This model is a fine-tuned version of Google's Gemma 4 and is not an official Google product.*
Exhibition at 'GENIAC' Hosted by METI and NEDO
This model will be demonstrated at the 'GENIAC' achievement announcement caravan, a generative AI project led by the Ministry of Economy, Trade and Industry (METI) and NEDO. You can experience the actual operating speed and quality of answers.
* **Fukuoka Venue:** April 8, 2026 (Wed) 13:00-16:00 @ RESOLA Hall (Directly connected to Tenjin Station) * **Tokyo Venue:** April 14, 2026 (Tue) 13:00-17:00 @ Shibuya Stream Hall (Directly connected to Shibuya Station) * **Details:** https://geniac-prize.nedo.go.jp/#outreachEvent
Future Outlook: Social Implementation of Secure AI and Innovation
We have been working on PoCs and practical AI support at universities. Moving forward, we aim for a society where universities and governments can 'use AI with peace of mind.' We will promote innovation through the social implementation of this model. We will provide new value to the domestic and international career support field by helping build personalized AI incorporating data from each university and developing next-generation services combined with XR fields such as smart glasses.
About Careerbot Co., Ltd.
* **Company Name:** Careerbot Co., Ltd. * **Careerbot DX Labo:** 2F NTV Yotsuya Bldg. (wework), 5-3-23 Kojimachi, Chiyoda-ku, Tokyo * **Representative:** Koji Okazaki (Specially Appointed Professor at Tokyo Metropolitan Institute of Technology / Visiting Professor at The Open University of Japan) * **Business:** DX consulting for university career centers utilizing AI/XR * **Established:** March 23, 2021 * **Official Site:** https://careerbot.tokyo/
FACT BOX
- Source: PR TIMES
- Category: News