Vietnamese
Japanese
EnglishNews
Latest insights, products and updates from the Fujigo Software Solutions team — AI, automation and digital transformation.
Recent articles
LLMRouter: Unified infrastructure for LLM routers
The LLMRouter paper just hit #1 trending on HuggingFace with 2,350 upvotes, proposing a unified architecture for developing, evaluating, and deploying routers for large language models.
Read articleDiagram Design: 29 Editorial Diagram Types for Claude Code
Diagram Design provides 29 editorial diagram types as self-contained HTML + SVG for Claude Code — no shadows, no Mermaid-slop. Trending on GitHub with over 4,400 stars in a single day.
Read articleRhetoric Reward-Hack: When AI gets fooled by persuasive writing in peer review
New research from UMaryland reveals how academic papers use rhetorical techniques to deceive AI reviewers, raising major questions about the reliability of automated evaluation systems.
Read articleCo-Evolution in Agentic Systems: When AI self-evolves beyond human design
A paper with 116 upvotes on HuggingFace explores the self-evolution capability of AI agent systems, opening a new era of AI that doesn't need to be redesigned from scratch.
Read articleComBodied Agents: A new era of human-centric AI
A trending paper on HuggingFace with 173 upvotes introduces a new AI agent model that puts humans at the center, opening up a direction for more natural AI interaction.
Read articleMacro: Unified workspace for teams with shared AI memory
A trending repo on GitHub with 1800+ stars, Macro integrates email, chat, docs, tasks, agents, calls, and CRM into one platform with shared AI memory, all @-linked together.
Read articleAgency Agents: Turn Claude Code and Cursor into a team of AI specialists with 1 click
The #1 trending repo on GitHub today with 958 stars. Agency Agents provides a collection of specialized AI agents — from frontend developers to content creators — installable into Claude Code, Cursor, and Codex with just 1 click.
Read articleBDH-CQ: A 150M-parameter model breaks reasoning limits with latent reasoning
The #1 trending paper on HuggingFace today proves that a small 150M-parameter model can achieve state-of-the-art results on ARC-AGI-1 through recurrent latent reasoning and in-context learning.
Read articleSemantica: Graph-native infrastructure for transparent and accountable AI agents
Trending #2 on GitHub today with 893 stars. Semantica provides graph-native infrastructure for AI agents — self-hostable, auditable, zero vendor lock-in. Dubbed 'The Open Source Palantir for AI agents.'
Read article
