Future Tech & Robotics

Mosaico Raises €7.5M Seed to Build the Data Infrastructure Behind Physical AI and Robotics

By mohitdeep1122 • • 7 min read • 📷 3 High-Res Visuals
Share Dispatch
𝕏 Post in Share Y Hacker News
⚡ TL;DR — Executive Summary

The €7.5 million is the new seed round, not the company's lifetime funding total. Including its earlier €1 million pre-seed, Mosaico has raised approximately €8.5 million. Its valuation has not been disclosed.

Italian startup Mosaico raises €7.5 million in seed funding led by United Ventures to build data infrastructure for robotics and Physical AI.
Italian startup Mosaico raises €7.5 million in seed funding led by United Ventures to build data infrastructure for robotics and Physical AI. 🔍 Click to zoom

Italian deep-tech startup Mosaico has raised €7.5 million (approximately US$8.3 million) in seed funding to solve a critical infrastructure problem holding back intelligent robotics: managing the enormous volumes of data generated by machines and their sensors. Led by United Ventures, with participation from OpenOcean and Connect Ventures, the investment will support the company’s open-source platform, international developer community and planned expansion into the United States. Founded in 2023, Mosaico is building the data layer that helps robotics engineers turn raw sensor recordings into searchable, reusable and structured information for developing and improving AI-powered machines.

Mosaico's open-source platform helps robotics engineers collect, organise, search and analyse complex sensor data from intelligent machines.
Mosaico's open-source platform helps robotics engineers collect, organise, search and analyse complex sensor data from intelligent machines.🔍 Click image to enlarge

A €7.5M Seed Round Backs Mosaico’s Physical AI Infrastructure Ambition

Mosaico announced its €7.5 million seed financing on November 16, 2026. The round was led by Italian venture capital firm United Ventures, with participation from OpenOcean and existing investor Connect Ventures. The company had previously secured €1 million in pre-seed funding in February 2026, also led by Connect Ventures. The latest investment brings its disclosed cumulative funding to approximately €8.5 million. The new capital will support further product development, international expansion of its open-source distribution and entry into the US market. Based in Parma, Italy, Mosaico was established by engineers who previously worked on autonomous-vehicle systems and sensor-data infrastructure. The financing comes at a time when companies are investing heavily in robotics, autonomous machines and Physical AI, increasing the need for software capable of handling the data generated by these systems. Mosaico’s strategy is to build an infrastructure product that can serve multiple robotics manufacturers rather than developing its own robot or targeting one narrow industrial application.

Robotics Has a Data Bottleneck That Slows Down AI Development

Modern robots depend on information collected from cameras, lidar, radar, GPS, inertial measurement units and other sensors. During testing, a single autonomous vehicle or industrial robot can generate enormous quantities of recordings containing images, movement measurements, environmental information and machine activity. Engineers need to organise this material, identify relevant events, investigate failures and prepare suitable datasets for training and evaluating AI models. The problem is that much of this work is still handled through fragmented software, manually organised folders and custom-built internal pipelines. Finding a short recording of a vehicle braking, a robot missing an object or a sensor behaving unexpectedly can require engineers to search through large volumes of data. Mosaico identifies this manual data-management process as a major source of engineering inefficiency. The company has argued that robotics teams can spend a substantial share of their time searching and organising sensor recordings instead of improving the machines themselves. As Physical AI expands, this problem is becoming an increasingly important part of the technology stack.

Mosaico Builds an Open-Source Data Engine for Intelligent Machines

Mosaico’s central product is an open-source, modular data platform designed specifically for robotics and Physical AI applications. It provides software for ingesting, storing, indexing, retrieving and querying information generated by different sensors. Rather than forcing engineering teams to build independent data pipelines for every project, the platform creates a common layer for managing robotics recordings across applications. Its architecture combines an SDK, which developers use to interact with the data, and a background service called mosaicod, which manages storage, indexing and retrieval. Mosaico is designed to work with existing databases and object-storage systems instead of requiring customers to replace their entire infrastructure. It also supports established robotics data formats, including ROS, ROS2 and MCAP. This architecture is important because robotics developers use different hardware, middleware and software frameworks. A flexible data layer can make it easier to reuse engineering tools across projects without tying every workflow to a single vendor.

Mosaico is building the data infrastructure that enables autonomous robots and intelligent machines to learn from real-world sensor information.
Mosaico is building the data infrastructure that enables autonomous robots and intelligent machines to learn from real-world sensor information.🔍 Click image to enlarge

Multimodal Search Helps Engineers Find Specific Events Inside Massive Sensor Datasets

One of Mosaico’s most important capabilities is its approach to searching complex sensor recordings. Engineers need more than a file-storage system: they need to find specific events across information collected by multiple sensors at different sampling rates. The platform supports structured queries across session metadata, topic names and physical sensor measurements, while its multimodal search capabilities are designed to make complex scenarios easier to locate through straightforward text-based requests. For example, a robotics engineer investigating an autonomous vehicle could look for recordings where a vehicle stopped at a pedestrian crossing, then retrieve the corresponding camera, GPS and movement information. The platform also provides precise timestamps so that developers can retrieve and replay the relevant section of a recording. Its synchronized replay capability brings information from different sensors together chronologically, making it easier to understand what happened before, during and after an event. This can reduce the time required to investigate failures and prepare useful examples for model development.

Data Lineage and Reusable Pipelines Could Improve Robot Reliability

Organising sensor recordings is only the beginning of the problem. Robotics teams must also understand how raw information changes as it moves through preprocessing, filtering, annotation, segmentation and AI training workflows. Without reliable tracking, developers can struggle to identify which original recording produced a particular training dataset or whether a transformation introduced an error. Mosaico addresses this through data-lineage capabilities that track transformations from source recordings to downstream datasets. Its platform also includes orchestration tools designed to coordinate third-party tools and automate activities such as labelling, segmentation and inference pipelines. Another part of the product is its focus on certifiable data pipelines, which can help organisations establish greater traceability across engineering workflows. These capabilities matter particularly for industrial robotics, autonomous vehicles and safety-sensitive systems, where teams need to understand how data was collected and processed before relying on the resulting AI model. Mosaico is positioning data management as part of the reliability infrastructure behind physical machines, rather than simply a storage or search function.

Open Source and On-Premises Deployment Are Central to Its Strategy

Mosaico has chosen an open-source model for the foundation of its platform, with its core software released under the Apache 2.0 licence. Customers can deploy the technology within their own infrastructure, giving engineering teams control over their data, applications and operating environment. This is relevant for robotics developers handling proprietary recordings, industrial information or sensitive autonomous-system datasets that they may not want to transfer to external cloud services. The company also offers a commercial enterprise model with additional capabilities such as auditing, dashboards, fine-grained access controls, user management, custom database connectors and pipeline certification. According to Mosaico, its open-source community has grown to more than 1,200 members since December 2025, including engineers and researchers associated with organisations such as Google DeepMind, Amazon, Zoox, Hexagon, ETH Zürich, EPFL and Politecnico di Milano. These are community affiliations rather than endorsements or customer contracts. The company’s strategy is to build adoption among developers first and establish a widely used infrastructure layer before the Physical AI market reaches broader commercial maturity.

Founders Bring Autonomous Driving Experience From Ambarella and Magneti Marelli

Mosaico was founded by Gabriele Lini, Francesco Di Corato and Federico Cabassi, who combine experience in autonomous-vehicle engineering, embedded systems and research. Lini serves as chief executive, Di Corato leads scientific development and Cabassi is chief technology officer. The founders’ professional backgrounds include Ambarella and Magneti Marelli, alongside academic and research experience connected to SISSA, the University of Pisa and the University of Parma. Their experience building autonomous-driving systems exposed them to the difficulties of organising large volumes of multimodal sensor recordings. Rather than developing another complete autonomous-vehicle system, they decided to focus on the shared data infrastructure that robotics companies need underneath their products. That horizontal strategy could allow Mosaico to serve autonomous vehicles, industrial robots, drones, agricultural machines and other Physical AI applications using a common software foundation. The company’s planned US expansion will help it reach a wider developer and robotics ecosystem, while the new funding will support continued engineering and community growth.

Mosaico’s Real Opportunity Is Becoming the Data Layer for Physical AI

The long-term opportunity for Mosaico is closely connected to the transition from software-based AI into machines that interact with the physical world. Robots will need to collect information, interpret their surroundings, learn from previous operations and improve their behaviour through repeated development and testing. That process creates an infrastructure requirement extending beyond chips, sensors and AI models: companies need dependable systems for managing the information that makes machine intelligence possible. Mosaico is attempting to occupy that layer with an open-source platform that can operate across different hardware and software environments. However, the market is still developing, and the company has acknowledged that Physical AI is not yet ready for widespread software purchasing at scale. It says early commercial contracts are progressing and enterprise deployments are underway, but the next test will be converting developer adoption into repeatable business demand. Its €7.5 million seed round gives it resources to pursue that transition. If Mosaico can become a standard way for robotics developers to manage sensor data, it could become an important infrastructure provider for the next generation of intelligent machines.

📷 Complete Article Visual Gallery (3 Pictures) Click any photo for full-screen view
Italian startup Mosaico raises €7.5 million in seed funding led by United Ventures to build data infrastructure for robotics and Physical AI.
Italian startup Mosaico raises €7.5 million in seed funding led by United Ventures to build data infrastructure for robotics and Physical AI.
Mosaico's open-source platform helps robotics engineers collect, organise, search and analyse complex sensor data from intelligent machines.
Mosaico's open-source platform helps robotics engineers collect, organise, search and analyse complex sensor data from intelligent machines.
Mosaico is building the data infrastructure that enables autonomous robots and intelligent machines to learn from real-world sensor information.
Mosaico is building the data infrastructure that enables autonomous robots and intelligent machines to learn from real-world sensor information.
mohitdeep1122 • TechFront360 Editorial Desk

Published by TechFront360.com following our strict fact-checking, benchmark verification, and independent technology journalism standards.

More in Future Tech & Robotics