1. Research and Development
  2. Fujitsu Kozuchi & Takane Tech Library
  3. Fujitsu Kozuchi technologies

Fujitsu Kozuchi

AI core engine, AI Trust, Optimization

Fujitsu Kozuchi, comprising approximately 100 proprietary AI technologies, and Takane, an enterprise generative AI specialized for business use, support diverse industries and business operations with domain-specific technologies and trusted AI.

AI Core Engines Tailored to Specific Industries and Business Needs

A transparent sphere engraved with a world map rests on the grass.

Materials Discovery

Fast AI-based material properties calculation and causal analysis of computed results can significantly reduce the search time for new materials.

Stacked documents with printed English text have their top pages turned over.

Terms and Conditions Analysis

Extract rules to be checked from terms and conditions documents reducing control departments’ workloads and preventing rule violations by user departments.

A transparent double-helix DNA molecule floats against a blue background.

Genomics AI

Achieve personalized medicine through AI-driven genetic analysis and high-performance computing.

Various types of screws and nuts are neatly arranged in the grid-like recesses of a white tray.

CAD Part Identification AI

Automatic CAD component part number identification even with unknown part numbers.

AI Trust: Enhancing Security and Building Trust in AI (AI Ethics, Security)

A yellow ring with a yellow sphere inside is set against a blue background, surrounded by scattered light blue spheres.

Cyber Attack Analysis and Countermeasure

Comprehensively identifies viable attack scenarios against the system and proposes optimal defensive measures.

An orange sphere is placed on a blue, wavy-shaped platform against a blue background.

LLM Vulnerability Scanner and Guardrails

Comprehensive assessment of LLM vulnerabilities and automatic application of optimal defensive techniques to Generative AI on LLMs.

Against an orange background, a large blue sphere and a small pink sphere sit on opposite ends of a balanced beam.

Fujitsu LLM Bias Diagnosis

Diagnose bias in LLMs from various points of views, allowing users to choose the most appropriate LLM.

Against a yellow background, three rolls of colored tape sit at the ends of matching strips laid flat.

SecureRAG for Data Poisoning Prevention

Making RAG-integrated LLM applications more secured and accurate.

Against a white background, a single red file stands out among a row of many light blue files.

SecureRAG for Safe Data Utilization

Creates easy-to-understand organization and visualization of past project documents as basis for current projects.

A magnifying glass focuses on and enlarges a single red "1" within a grid of 0s and 1s.

SecureRAG for Data Loss Prevention

Securely leverages organizational data with RAG and LLMs, while preventing the leakage and misuse of confidential information.

Optimization

In an office, one man operates a keyboard while another points to a thermal map on a monitor.

AI Surrogate Model Trial Platform

Reduce the time and cost of CAE analysis via the easy introduction of surrogate models.

A red square component sits on a black circuit board, which features intricate red circuit paths.

Fujitsu GPU-less AI inference

Using Fujitsu's Arm CPU optimization for faster, more efficient AI service realization, removing the need for GPUs.

Connect with Fujitsu Research

LinkedIn

X

YouTube

page top