goNEON Agentic Systems Secures €160K to Accelerate AI-Powered Infrastructure Planning

A spin‑off from ETH Zurich is betting that infrastructure planning, an industry still built around manual engineering workflows and disconnected software, is overdue for the same kind of automation reshaping other technical fields.

goNEON Agentic Systems has secured 160,000 euros, roughly 150,000 Swiss francs, in funding from Venture Kick to accelerate development of its agentic AI platform for infrastructure planning. The company was incorporated in February of this year and is headquartered in Zurich.

The problem goNEON is targeting is a familiar one to anyone who has worked in urban planning, transport engineering, or infrastructure development. Evaluating even a handful of design alternatives for a road network, utility system, or transit line typically requires weeks or months of manual analysis, drawing on fragmented software tools and outdated models that struggle to capture how complex, city‑wide systems actually interact with one another. As infrastructure projects grow more complex and budgets tighten, the slow pace of traditional planning has become a genuine bottleneck rather than just an inconvenience.

goNEON's platform addresses that bottleneck with what the company describes as the first fully agentic infrastructure‑planning service. Rather than requiring planners to manually model every alternative, the system automatically generates infrastructure designs based on engineering requirements, local regulations, and real‑world physical constraints. Planners, engineering firms, and infrastructure operators can prompt the company's AI agent, which it calls N!, and receive technically feasible, regulation‑aware designs for roads, utilities, and transit networks in minutes rather than the years such analysis traditionally requires.

Importantly, the platform is built to support engineers rather than replace their judgment. Instead of generating a single final design and asking planners to accept or reject it, the system rapidly produces multiple technically feasible planning options, allowing teams to compare, assess, and refine alternatives in a fraction of the time traditional methods demand. The company says its models are also constraint‑aware, meaning the system understands physical and regulatory limits well enough that it cannot generate plans that violate network consistency or real‑world feasibility, a distinction the founders argue separates goNEON from more generic AI tools applied loosely to planning problems.

The company was founded by Raphael Eder, who serves as chief executive, and Dr Lukas Ballo, who serves as chief technology officer. The two combine backgrounds spanning entrepreneurship, artificial intelligence, and urban planning, with Ballo holding a PhD and master's degree from ETH Zurich and Eder holding degrees from UC San Diego, the London School of Economics, and ETH Zurich. Their approach draws directly on research conducted at ETH Zurich, where the founders developed computational methods for the evidence‑based redesign of urban transport systems before spinning the technology out into a standalone company.

goNEON describes its own funding journey candidly, noting that Venture Kick helped the company move from academic research toward actual paying customers. The program pushed the team to validate market demand early, connected it with a network of founders and investors, and provided the initial momentum needed to think about building a global business rather than a research project confined to one university.

With the new capital, goNEON plans to prioritize pilot projects with the strongest commercial potential, convert its most repeatable planning workflows into scalable software modules, and validate the specific use cases that will form the foundation of its broader infrastructure planning platform going forward. The company says it is already live in five countries, spanning Switzerland, Germany, Austria, the United Kingdom, and the United States, working across five distinct production use cases, including automated geometric verification of sight triangles at driveways and intersections for building‑permit workflows, a task the company plans to make available soon as an agent through the Microsoft marketplace.

goNEON is targeting what it describes as the 1.5 trillion dollar global architecture, engineering, and construction services market, positioning its agentic planning tools as a way to convert infrastructure planning from one‑off, project‑based consulting work into a repeatable, largely automated computational process. Given how heavily agentic AI investment has concentrated in software, customer service, and coding tools over the past two years, goNEON's early traction applying the same underlying approach to physical infrastructure planning marks one of the more concrete examples of agentic AI moving into a genuinely underserved, real‑world engineering discipline.