SuperAPI AI Directory
Partner with us中文AI newsFavorites

AI agents

2,963 results · Page 52 of 83

Kinaxis

AI agents

Kinaxis is a supply chain decision platform that provides end-to-end transparency through context-aware agentic AI, a purpose-built decision platform, and more than 40 years of supply chain experience, helping enterprises anticipate disruptions, coordinate responses, and accelerate decisions. Its core platform, Maestro, is based on concurrency technology, continuously synchronizing signals, plans, and decisions, and supports understanding impact, evaluating trade-offs, and adjusting in real time. The platform offers composable AI, demand sensing and forecasting, supply planning, inventory optimization, production scheduling, and other solutions, and also has Kinaxis Planning One for growing teams to quickly modernize planning. It is suitable for planning and decision-making scenarios in complex supply chain organizations.

Read more

KwaiAgents

AI agents

KwaiAgents is a collection of a series of Agent-related works open-sourced by the KwaiKEG team at Kuaishou Technology, including the lightweight information retrieval agent system KAgentSys-Lite, the KAgentLMs series of large language models with planning/reflection/tool-use capabilities, over 200,000 Agent-related instruction fine-tuning data KAgentInstruct, and over 3,000 manually edited automated evaluation data KAgentBench. The project builds a general-purpose information retrieval agent based on large language models, supports deploying model inference services via vLLM and FastChat, can also use llama-cpp-python to provide an OpenAI-compatible interface, and supports running agents driven by ChatGPT or local models.

Read more

l2mac

AI agents

L2MAC (Large Language Model Automatic Computer) is a general-purpose stored-program automatic computer framework based on LLMs, adopting the von Neumann architecture, for LLM multi-agent systems. It solves complex tasks by generating large amounts of consistent output, unconstrained by the LLM's fixed context window. The framework can generate an entire complex codebase or a whole book from a single-line prompt, achieving 90.2% Pass@1 on the HumanEval benchmark and ranking third on the global coding industry standard leaderboard. L2MAC was fully open-sourced on April 23, 2024, and was presented at ICLR 2024.

Read more

LabClaw

AI agents

LabClaw is a skills library for biomedical AI workflows, providing an operational layer for LabOS (the Stanford-Princeton AI joint scientist project) and supporting dry-lab reasoning, experimental protocol orchestration, and agent workflows, forming a closed loop with LabOS XR/wet-lab execution. It is not monolithic software; rather, it contains 240 production-ready SKILL.md files covering biology, lab automation, vision/XR, drug discovery, medicine, data science, literature research, and scientific visualization, and can be installed as a whole or selected by subfolder as needed. Each skill defines when an OpenClaw-compatible agent should use a tool, how to invoke it, and what kind of result to produce, making it suitable for researchers who want a modular skills layer.

Read more

Labelbox

AI agents

Labelbox is a platform that builds training environments and provides expert data for frontier AI labs, while also providing enterprises with AI agents (Recursion) that can work autonomously. Its agents can be dispatched to handle recurring or triggered work across software engineering, reliability operations, security engineering, machine learning research, evaluation, product, finance, sales, and other teams, such as fixing flaky tests, upgrading dependencies, investigating alerts, reviewing code security, running experiment sweeps, updating documentation, and refreshing financial forecasts. Each agent starts and runs autonomously in the background until the work is complete, delivering reviewable pull requests, preview links, or reports.

Read more