Think First, Then Act -- Sereact Cortex 2.0 Robotic Brain
A VLA + world model robotic brain that simulates candidate trajectories before executing. 287 upvotes on Product Hunt. Backed by a fresh $110M Series B.

A Robot That Imagines Before It Moves
287 upvotes on Product Hunt. Sereact just shipped v2 of its robotic brain, fresh off a $110M Series B (maker: Ralf Gulde).
Most robotic AI works reactively -- see something, do something. Cortex 2.0 flips that. Before the robot moves, it simulates multiple candidate trajectories, evaluates each one for safety and efficiency, and picks the best path. Like how you mentally rehearse grabbing something awkwardly shaped before your hand actually moves.
What It Does
Cortex 2.0 simulating multiple candidate trajectories simultaneously
Cortex 2.0 combines a VLA (Vision-Language-Action) model with a world model.
- VLA: Understands visual input from cameras and natural language commands, converts them into robot actions
- World Model: Learns physics rules to predict "if I do this, what happens next"
Together, the robot can predict the outcome of actions and choose the best one before moving.
First Impressions
The demo video on Product Hunt is compelling. In a warehouse environment, the robot picks various oddly shaped objects, and before each pick you can see it simulating 2-3 candidate trajectories before selecting the optimal path. Sereact claims a 40%+ reduction in failure rate compared to v1.
Three Key Features
1. Trajectory Simulation. Simulates multiple candidate paths before acting. Calculates collision probability, success rate, and time for each, then picks the winner.
2. Zero-Shot Adaptation. Handles new objects without retraining. The VLA model infers shape, material, and weight from vision and language understanding.
3. Natural Language Commands. Understands and executes plain-language instructions like "move the red box to the left conveyor."
Pricing
Paid, enterprise-only. Exact pricing undisclosed. With $110M in fresh funding, they're in enterprise sales mode.
Who Benefits
Logistics, manufacturing, and warehouse operations teams looking at robotic automation -- especially environments with high product variety where items change frequently. Traditional industrial robots need reprogramming for every new item. Cortex 2.0 doesn't.
Similar Tools
Cortex 2.0 in a live warehouse environment
- NVIDIA Isaac Platform: Broader robot simulation and training platform. More versatile but more complex to set up.
- Covariant Brain AI: Similar robotic picking AI. Direct competitor.
- Google DeepMind RT-2: Research-stage VLA model. Not yet commercialized.
Robots that think before they act. That's the pitch, and the demo backs it up.
References
관련 기사

100x Less Energy, 3x More Accurate: The Neuro-Symbolic AI Breakthrough
Tufts researchers combined neural networks with symbolic reasoning to cut AI energy use by 100x while boosting accuracy from 34% to 95% in robotic planning tasks.

Yann LeCun's AMI Raises $1.03B Seed — The Biggest Bet Against LLMs
Deep learning pioneer Yann LeCun launches AMI Labs with a record $1.03B seed round. Full breakdown of world models vs LLMs, JEPA architecture, investor roster, competitive landscape, and what this means for the future of AI.

The Robotics Mega-Round Era: $1.2B+ Deployed in a Single Week
Mind Robotics $500M, Rhoda AI $450M, Sunday $165M, Oxa $103M -- AI has left the screen and entered the physical world
AI 트렌드를 앞서가세요
매일 아침, 엄선된 AI 뉴스를 받아보세요. 스팸 없음. 언제든 구독 취소.
