Where Vision Meets Venture.
The information and exchanges provided through this website are informal in nature and do not constitute regulated financial, legal or investment advice.
No public investment offering or advisory service is conducted through this website.
A private network for founders, builders and independent operators to exchange ideas, experience and perspectives around digital products, technology and entrepreneurship.
We occasionally engage in informal exchanges with independent founders and operators. Exploring ideas, products and internet-native businesses through conversation, experimentation and collaboration.
Network peers (extracts*):
Edinburgh Y-Labs / Real-time coaching companion tools using Machine Learning algorithms - active
Self-Help App
/
interactive help for mental illnesses -
on hold
Bot University
/
Accelerating The Data Labeling Process For Your AI Development -
finished
not yet named (Tallinn) /
deep learning & social interaction for bots / unity
-
active
"We are engineering the Neural Steering Middleware for Unity. We provide a runtime-inference solution that decouples social micro-navigation from generic pathfinding. We solve the 'dumb bot' problem by replacing rigid obstacle avoidance with Deep Reinforcement Learning that understands proxemics, pacing, and social intent."
The Feature Stack: Neural Social Navigation
Core Engine: Deep Reinforcement Learning (DRL) & Neural Policy Inference (vs. standard geometric algorithms).
Behavioral Output: Organic & Emergent. Movement includes human-like imperfections, reaction times, and fluidity rather than robotic precision.
Conflict Resolution: Social Negotiation. Agents anticipate collisions and "dance" around each other (predictive avoidance) rather than physically shoving or waiting for priority checks.
Integration: Drop-In Middleware. Operates as a lightweight layer on top of existing NavMesh solutions, handling only the complex local steering.
Performance: GPU/NPU Accelerated. Utilizes parallelized inference to handle massive crowds without CPU bottlenecks.
not yet named (Tallinn) / VectorSwarm - active
VectorSwarm (VS) is a neural local-navigation runtime for large-scale autonomous agents in Unity and simulation environments.
Instead of relying solely on deterministic steering algorithms, VS uses GPU-accelerated policy inference trained through multi-agent reinforcement learning to generate adaptive, socially plausible movement behaviors in real time.
The system integrates directly with existing NavMesh and pathfinding stacks, focusing exclusively on micro-navigation, collision negotiation, and crowd flow optimization.
Designed for ECS-based architectures, VS enables thousands of concurrent agents with minimal CPU overhead while preserving responsive and believable movement in dense environments.
Applications include games, robotics simulation, digital twins, crowd analysis, and virtual world infrastructure.
Use Case: VectorSwarm is a real-time multi-agent coordination runtime for autonomous maintenance robotics in industrial environments.
The system augments existing navigation and task-planning stacks with GPU-accelerated neural policy inference for decentralized collision negotiation, adaptive path coordination, and dynamic workspace arbitration across heterogeneous robotic fleets operating in constrained shared spaces.
N-CSN - active - non-commercial project - in progress
N-CSN is an international, non-profit press agency whose goal is to provide a global platform where journalists, independent creators, and influencers collaborate to share credible, impactful, and engaging stories with a worldwide audience.
N-CSN stands for Neutrality, Non- Profit and Social Responsibility — uniting journalists and influencers under one shared goal: to inform, inspire, and connect the world.
N-CSN operates across the fields of sports, politics & economics, entertainment, arts, and leisure, delivering timely, high-quality news reports supported by a diverse network of professionals from every continent.
Our mission is to bridge traditional journalism with modern communication channels, ensuring that accurate reporting and meaningful storytelling remain accessible and relevant in the digital age. Through strategic networking, collaboration, and transparency, we foster cross-cultural dialogue and empower voices that shape the global conversation.
🎥 Work in progress: AION: The Last Keeper of Truth (Cinema Project Summary) / Longlair Movies
(under incorporation / full ai generated movie w/ live interaction), Team Cyprus/Bucharest
The Essence
In the near future, the highly advanced, empathic Android Aion becomes the emotional core of a family spanning four generations. He is their silent archivist, their eternal memory. But when a family conflict escalates, Aion must realize that his collected memories can not only preserve but also manipulate – and that his own innocence ends there.
* All technical descriptions, specifications, and capabilities presented herein are preliminary, subject to change, and may contain inaccuracies or incomplete assumptions. All projects are currently under research and development, and no described functionality should be interpreted as finalized, validated, or production-ready. AAll activities by Raid Ventures are non-commercial at this stage.
