"thtRobotics is building the most versatile packing system - ever."
Grocery, Meal kits and Food Fulfillment are Breaking Under Manual Labor
Manual order packing drives a significant share of fulfillment costs. For baskets under €65, groceries operate at a loss.
People quit within 3-4 months, leading to 100%+ annual turnover and high HR costs (€10-15k per replacement).
1-3% mis-pick rates cost sites ~€0.5M annually and lead to significant customer churn.
A complete, adaptive solution to automate manual fulfillment.
Patented adaptive gripper technology that handles any fresh food item with human-like dexterity.
Our in-house AI-native vision sees through clutter and recognizes thousands of SKUs instantly.
Intelligent motion planning that enables fast deployment, high speed, and simple scalability.
Overview of the Noyce AI Pick&Pack system in production.
Dexter Gripper handling mixed grocery items.
DeepEye vision system identifying cluttered SKUs.
Noyce Brain orchestrating fast, adaptive motion plans.
€10K installation fee + €6K-8K/month subscription per robot cell.
Includes maintenance, software license & updates, and performance monitoring.
ROI ~ 2 months. Low upfront cost. Cash-flow friendly. Scales fast.
€100K-120K per robotic system + €15K annual license.
Ideal for enterprises with CapEx preference.
ROI ~ 1 year. Full ownership. Long-term control.
2 contracts signed, 3rd soon + 5 in making. Robot deployed at Europe's largest online grocer!
We partner with system integrators to accelerate worldwide deployment.
Manual fulfillment labor is scaling unsustainably. We unlock efficiency at a global scale.
The €10B+ spent annually on manual picking labor is projected to reach €40-50B by 2030. We estimate a €24B market for robotics to automate this work.
We're at the crossroads of three powerful forces driving a once-in-a-generation market shift.
The existing model of manual labor in food fulfillment is economically unsustainable. Acute labor shortages, soaring costs, and razor-thin margins have created an urgent, existential need for a new solution.
A recent, powerful convergence of AI-native perception (seeing in clutter) and adaptive gripping (human-like touch) has unlocked true dexterity for machines. This was technically impossible at scale just five years ago.
The industry is stuck building robots for perfect, structured worlds. We are breaking that pattern by building robots for the world as it is. This unlocks the largest, most valuable markets—like fresh food—that old automation can't touch.
Roboticists, AI engineers, mathematicians, and builders blending deep tech with startup grit.
founder & CEO / CTO
BSc Mechanical Engineering (UN) / Msc Mechatronics & Control (UN)/ PhD roboticist (UN) / 17 academic awards / 14 years building hardware-AI stacks / 43 published papers / leading multi-institution EU projects/teams; invented the adaptive-gripper technology and raised €2.4 M to launch thtRobotics; ex-founder of Tywai, ex-researcher at FMN/TU Berlin/TU Ilmenau; ex-post-doc at LUT/UiO; ex-associate professor at OsloMet -now steering the mission to automate grocery-handling at scale.
co-founder, COO & Head of Sales
10-year hardware rain-maker who led deals with Microsoft, Intel, Samsung & Sharp. Ex-GM at Eve Devices, former Account Executive (Senders) and consultant (Nametage); now driving global revenue for thtRobotics adaptive-robotics line.
Senior Robotics Engineer
BSc Aerospace Eng (Cairo) / MSc Robotics & Computer Eng (Tartu) / PhD Robotics, Politecnico di Milano, where he authored RL-driven cloth-manipulation robots (4 publications) and designed the first single-arm laundry-loading system. Former CAE engineer at NorthWind and robotics intern at DLR; now channeling 10 yrs of robot path planning R&D into thtRobotics adaptive food-handling line.
Robotics Engineer
OsloMet B.Sc. (electronics & automation, top grades) with 4 yrs PLC/ROS chops. At thtRobotics he runs end-to-end Dexter gripper development and production-machining. assembly and iterative upgrades - while programming motion paths that keep our picking cells on-spec and on-time.
AI Computer Vision Engineer
Dual-MSc mathematician (AI + Statistics) who turned C++ rigor into production-grade vision. Architect behind DeepEye - thtRobotics computer-vision stack - she designs CUDA/ROS pipelines, trains the core AI models, and has already cut deployment time 3x- while boosting detection accuracy and throughput for our adaptive-gripping cells.
AI Computer Vision Engineer
Built thtRobotics first DeepEye vision stack over 2.5 yrs: trained synthetic-data CV models for various SKUs and authored the full front-/backend GUI that runs every robot cell. Blends vision science with product UX to turn adaptive picking into a one-click operator experience.
AI Computer Vision Engineer
Math-and-physics Olympiad medalist turned AI specialist. ETF Belgrade Signals & Systems graduate, ex-Asseco enterprise developer. Now designing the intelligent perception that lets thtRobotics robots see, understand and grasp in real time.
We are raising a €1.2M late seed round to reach commercial scale.
This round enables us to prove commercial viability with minimal dilution, positioning for a strong Series A at a significantly higher valuation.
Contact Us