Addness Inc. (Headquarters: Yotsuya, Shinjuku-ku, Tokyo; CEO: Kota Mikami), a global leader in generative AI innovation, is pleased to announce that it has obtained a patent (Patent No. 7820002) for an information processing system that allows AI to automatically optimize and update roadmaps based on a learner's progress and points of difficulty.

Background of the Patent The challenges of "roadmap obsolescence" and "instructional costs" in educational settings. In traditional human resource development and educational schools, the roadmaps provided to students are rarely updated after creation. This has frequently led to cases where students drop out without anyone visualizing exactly where they are struggling.

While existing project management tools can "visualize" progress, responding to students who hit a wall still relies on the experience and judgment of instructors. This creates a structural limitation where maintaining educational quality and reducing instructional costs become a trade-off. Furthermore, even when attempting to utilize generative AI, methods that process vast amounts of information as-is have lacked sufficient accuracy, making it difficult to generate practical, usable roadmaps.

Overview of the Patented Technology "Detecting" stumbling blocks and automatically generating AI "detour routes." The newly patented technology is a system that treats learner inquiries as "signals of difficulty," allowing AI to continuously improve the roadmap to the goal.

<Unique and Innovative Points of this Patent> 1. Dynamic Branch Route (Detour Route) Generation The system automatically processes user questions and dynamically adds or updates "supplementary steps" or "detour routes" in the roadmap. This creates an environment where students can continue learning without coming to a standstill.

2. High-Precision "Two-Stage AI Prompt Processing" When utilizing generative AI, the system employs a two-stage process: first, it structures the information (first prompt), and then it generates support information based on that structured data (second prompt). This enables the generation of roadmaps and answers with significantly higher precision than conventional AI generation.

Value Provided by this Technology Reducing the burden on instructors and maximizing student achievement rates. With the introduction of this technology, it becomes clear to students what they need to do next, and the time instructors spend on individual support is significantly reduced. Because past inquiries and answers are accumulated and shared within the roadmap, there is no need to respond to duplicate questions. As a result, educational institutions can expect a decrease in student dropout rates and an increase in goal completion rates, enabling the provision of a high-quality educational environment that does not rely on human-dependent instruction.

Future Outlook Becoming a social infrastructure that supports all "goal achievement." Using this patented technology as a foothold, Addness Inc. will accelerate its expansion into all scenarios where "people achieve goals," including not only educational schools but also corporate employee training and onboarding. By providing a mechanism where AI stays close to an individual's vision and goals, accompanying and supporting their actions as a social foundation, we aim to build a society where everyone can realize what they want to do without experiencing frustration.

<Patent Overview> Invention Title: Information Processing System, Information Processing Method, and Program Patent Number: Patent No. 7820002 Registration Date: February 16, 2026 Patent Holder: Addness Inc. Inventor: Kota Mikami

[Company Overview] Company Name: Addness Inc. Headquarters: Yotsuya Tosei Building 5F, 4-3-50 Yotsuya, Shinjuku-ku, Tokyo Representative: Kota Mikami, CEO Established: July 26, 2021 Business Description: Next-generation education business, Web marketing support business URL: https://skill.addness.co.jp/public22

FACT BOX

  • Source: PR TIMES
  • Category: research