AI Briefing#ai-agent
The latest AI and developer news about #ai-agent, with the original source and a short summary.
The latest AI and developer news about #ai-agent, with the original source and a short summary.
The feature allows the agent to run multiple tasks at once while users control login choices and approve downloads.
Achieved a 2x contract conversion rate by leveraging AI tools and using easy-to-understand language in consultations.
Banksalad has filed three patents for technologies that control errors in financial transactions by AI agents.
BankSalarad has been designated as an FSC Innovative Financial Service to support proxy applications for loan pre-approval.
The system uses isolated VMs and conversation-specific undo to let users direct file tasks while retaining control over permissions and changes.
Kakao has released the 3B-scale multimodal decision-making model Kanana-2-v-decide as open source.
Carrot launched Korea's first advertising MCP as a public beta on September 16.
MUSINSA has developed the DeciML and NeXA platforms to handle decision-making and discovery beyond prediction.
We confirmed cost savings by reducing system call overhead to runc levels through microVM isolation and implementing non-billing for idle CPU.
Harness size was halved due to improved model performance, but verification mechanisms must be maintained.
CJ Olive Networks secured both control and convenience for enterprise AI usage through an AWS-based LLM Gateway.
Ablely managed the lifecycle of AI agent workloads using Amazon EKS and Agent Sandbox.
OpenQodex is an open-source tool where a separate AI session reviews code changes before pushing, achieving a 98% reproduction rate for seeded bugs.
Figma shares how its latest unreleased editing model, post-trained on Qwen 3.8 27B, scores 77.8% on internal evals, surpassing the top third-party model, while the general availability agent sees 20% faster task completion.
CopilotKit has open-sourced OpenDuet, a self-hostable proactive work agent.
The article shares how PMs use AI to automate repetitive tasks, allowing them to focus on planning and judgment.
The Kakao Recommendation Team applied Data Catalog and Analyst Persona to unify AI analysis standards.
The system separates LLM interpretation from code-based calculation to ensure accurate daily reports for LINE OA and Display Ads.
A Mushinseok developer identified three judgment limits when working in parallel with AI agents.
The patch release resolves 61 pull requests to fix Gateway update checks, preserve agent conversation context, and correct macOS Talk and Bun startup bugs.
The agent uses a separate safety classifier to block sensitive data and requires user approval for high-risk actions like payments and deletions.
IHF 2026 highlighted the need for data foundations and operational governance through case studies from MGB and Narayana Health
Wanted built a system to manage neglected app reviews in one week by leveraging internal tools and AI.
An API for agents and RAG that charges only when an HTTP 200 response is returned along with actual results.