We spend a lot of time thinking about what the right "sweetspot" for low-code/no-code agent building is I don't think it's fully no code (we see that most agents still involve a substantial amount of custom code) But there is DEFINITELY a visual process to building agents
AGENTS
-
OmniParser: Vision-Based GUI Agent by Microsoft
By
–
📷OmniParser: Vision Based GUI Agent
— Satya Mallick (@LearnOpenCV) 18 mars 2025
Microsoft's OmniParser represents a breakthrough in how AI systems interact with graphical interfaces. By using a pure vision-based approach to extract structured elements from UI screenshots, OmniParser helps large multimodal models like… pic.twitter.com/3qBzaxWV3IOmniParser: Vision Based GUI Agent Microsoft's OmniParser represents a breakthrough in how AI systems interact with graphical interfaces. By using a pure vision-based approach to extract structured elements from UI screenshots, OmniParser helps large multimodal models like
-
Human in Loop Support for Parallel Multi-Agent Systems
By
–
Want to have human in the loop support for multiple agents running in parallel? Just got a lot easier
-
Evaluating and Scaling AI Agents with Writer CTO
By
–
Evaluating and Scaling AI Agents Writer is a full-stack generative AI platform that delivers transformative ROI for the world’s leading enterprises How do they do it? I'm going to be chatting with their CTO Waseem Alshikh this Thursday in SF https://
lu.ma/x7ohaj2c -
Arcee Conductor: Intelligent Model Routing with $200 Free Credits
By
–
Announcing Arcee Conductor with $200 FREE Credits —
Arcee Conductor is a new standard for intelligent model routing. It intelligently routes your prompt to the best model, to efficiently deliver precise results, for any task. Our pioneering work in small language models -
Zoom’s AI Companion Becomes Agentic with Task Automation
By
–
Zoom's AI Companion is going agentic with new tools to identify and complete tasks across the platform It is getting new memory and reasoning smarts to leverage the right agentic tools for each task, like executing action items or scheduling follow-ups
-
MindVLA Navigates Complex Traffic Using RL and Multi-Agent Modeling
By
–
MindVLA can also navigate complex traffic environments by jointly modeling its own vehicle behavior and the trajectories of other vehicles. It employs RL to learn from mistakes, continuously optimizing its decision-making.
-
MindVLA: Fast-Slow Thinking with Action Token Diffusion
By
–
MindVLA integrates fast and slow thinking modes and can switch between them autonomously. When deployed, it uses Diffusion to decode Action Tokens into optimized trajectories.
-
KG-Powered Multi-Agent System Reduces AI Hallucinations
By
–
Using a KG-powered multi-agent system to generate accurate biographies? That's exactly what a team from Zhejiang University just released. They built a historian AI called AIstorian, which has a much lower hallucination rate and significantly higher factual accuracy.
-
OpenAI Swarm and Pydantic AI: Abstraction vs Code Clarity
By
–
But don’t OpenAI swarm and pydantic ai just abstract away code? And obfuscate just as much