Sándor Felber, born in Törökbálint and still in his twenties, studied in Edinburgh and at the University of California, Los Angeles (UCLA). He worked on autonomous race cars and later contributed to actuator control for the Tesla Optimus project. He led a small team at the MIT CSAIL lab focusing on AI-based control for humanoid robots, participated in the South Park Commons Founder Fellowship, and ultimately co-founded Minerva Humanoids with Maurice Rahme, the former technical lead of Boston Dynamics' Stretch Manipulation team.
Record pre-seed funding
According to the report, the company—founded in autumn 2025—rapidly attracted investor interest and closed a nearly $10 million (about three billion forints) pre-seed financing round this spring, led by the US venture firm General Catalyst. This sum is described as a record for Hungarian-founded startups at such an early stage. The firm had previously completed smaller, closed rounds; a Hungarian investor joined only in this round: Bálint Pásztor, founder of DiffuseDrive, invested as an angel.
Mission: replacing hazardous work and cutting climate impact
Felber says roughly 7,000 people a year die in the energy sector alone while performing jobs that humanoid robots could potentially do. Minerva’s mission is to transfer dangerous, human-risking tasks to robots where feasible and also to reduce the environmental footprint of client industries—for example by replacing helicopter trips with VR-enabled remote interventions.
Three development tracks
Minerva's R&D splits into three primary streams:
- Minerva intelligence: an application-specific AI system designed to help the robot perceive its immediate environment, adapt to it and perform assigned tasks. This is not a general large language model (LLM) but a task-focused model.
- Hardware design: building the robot body with task-specific considerations (for instance, climbing icy stairs while holding a railing); Minerva is developing its own hardware architecture.
- Remote expert solutions: robots serve as the field-executing ‘hands’ for remote specialists. A distant expert dons a VR headset, connects to the robot on site, guides it to the right spot and issues instructions to complete hazardous repairs or maintenance.
Use cases and demonstrations
Target sectors include energy (notably oil and gas), public safety and defense (e.g., bomb disposal and CBRN tasks). The company presented a prototype at a Slovak conference this spring in front of NATO bomb-disposal experts. Felber stresses that for complex jobs human oversight will likely remain necessary in the foreseeable future, while robots can already take over many repetitive and dangerous duties.
Product development, pilot and go-to-market plans
Minerva does not yet sell a broadly market-ready robot; the recent funding primarily supports R&D. The company plans a pilot project starting this fall to test prototypes and fix shortcomings, and aims to launch its first commercial robot next year.
Their development and production layout is mixed: hardware work is concentrated in Europe—mainly Germany—while software development is based in San Francisco, allowing them to leverage strengths from both regions. Felber also noted that the European Union’s detailed regulatory framework helps plan and audit development processes.
Business model and pricing
Minerva intends to offer robots via a subscription model (robot-as-a-service), with ongoing software and feature updates alongside the hardware. No exact prices have been published yet; Felber estimates robots will likely cost in the several-hundred-thousand-dollar range depending on configuration. He also notes that for some projects firms care less about sticker price when the labor simply cannot be sourced.
Volumes and timeline
The immediate plan is to build and test a few prototypes this year, scale to the order of hundreds of units next year, and potentially reach several thousand deployed robots within a few years. The company does not plan classic mass manufacturing because maintaining strict quality checks for each unit would be difficult at very high volumes.
Risks and opportunities
Investors are betting on the team and on the underlying technology. In the broader robotics landscape the size of this round is not unprecedented, and Felber acknowledges that LLM-focused startups operate under different investment dynamics given their broader applicability. Supporters highlight Minerva’s focus on task-specific humanoids as an advantage relative to teams building generalist humanoid platforms.
Overall, Minerva Humanoids aims to reduce human risk in hazardous jobs—potentially saving thousands of lives annually—and to lower environmental impacts in heavy industries, while advancing through pilot projects toward commercial deployments.
