Yumaniq

Cover Block

PUBLIC

Attribute	Value
Name	Yumaniq
Tagline	Motor Intelligence Infrastructure for Physical AI, enabling intelligent machines to learn complex manipulation skills.
Headquarters	Israel
Stage	Pre-Seed
Business Model	SaaS
Industry	Deeptech
Technology	Robotics
Growth Profile	Venture Scale
Founding Team	Solo Founder (Nitsan Sharon)

Links

PUBLIC

• Website: https://www.yumaniq.com
• LinkedIn: https://www.linkedin.com/in/nitsan

Executive Summary

PUBLIC Yumaniq is building motor intelligence infrastructure for physical AI, an early bet on the software stack that will enable robots to learn and adapt in unstructured environments. The company's proposition deserves investor attention as a foundational play in a sector where data management and model deployment remain fragmented, despite rapid advances in AI models [yumaniq.com, retrieved 2024]. Founded by Nitsan Sharon, a former Oracle and Amdocs technology executive, the company is developing RAST, a software stack designed to convert expert demonstrations into executable motion policies while enforcing safety constraints [yumaniq.com, retrieved 2024] [LinkedIn, retrieved 2024]. Its differentiation hinges on providing a unified layer for data, training, and deployment, aiming to accelerate iteration for robotics teams. The company operates in stealth mode from Israel, with no public funding rounds, customer logos, or external press coverage yet available, placing it at a pre-seed stage of validation. The next 12-18 months will be critical for Yumaniq to secure initial capital, demonstrate its technology with a lighthouse customer, and expand its technical team beyond the solo founder structure.

Data Accuracy: YELLOW -- Product claims and founder background are sourced from the company website and LinkedIn; funding, traction, and market data are unconfirmed.

Taxonomy Snapshot

Axis	Value
Stage	Pre-Seed
Business Model	SaaS
Industry / Vertical	Deeptech
Technology Type	Robotics
Geography	Israel
Growth Profile	Venture Scale
Founding Team	Solo Founder

Company Overview

PUBLIC

Yumaniq presents as an early-stage deeptech venture, positioning itself at the intersection of robotics and artificial intelligence. The company is headquartered in Israel, a detail confirmed by both its website and the founder's public profile [yumaniq.com, retrieved 2024] [LinkedIn, retrieved 2024]. Its founding narrative centers on building foundational infrastructure for a new class of intelligent machines, a concept the company terms "Physical AI." The public record does not include a founding year or details of incorporation.

The company's founder, Nitsan Sharon, serves as both Founder and CTO. His professional background includes roles at Oracle, where he was CTO for the Strategic Clients Group in Israel, and at Amdocs [alanquayle.com, retrieved 2026] [RocketReach, retrieved 2024]. Sharon has also participated in public discussions as a climate tech professional, indicating a breadth of technical and industry perspective [linkedin.com/in/nitsan/, retrieved 2026]. The company's leadership structure appears lean, with no other co-founders or executives named in public materials.

Key operational milestones are not publicly documented. There are no press releases announcing a product launch, a first customer, or a strategic partnership. The company's primary public presence consists of its website and a technical blog, which outline its product vision and architectural approach. This suggests Yumaniq is operating in a quiet, development-focused phase typical of very early-stage hardware-software ventures.

Data Accuracy: YELLOW -- Company location and founder identity are confirmed; founding details and milestones are not publicly available.

Product and Technology

MIXED

Yumaniq's public positioning frames its offering as a foundational software layer for robotics teams, analogous to MLOps tooling but for physical motion. The company describes its product, RAST, as a software stack that runs alongside a robot's existing control systems to enable learning from demonstration [yumaniq.com, retrieved 2024]. According to the company's blog, this approach is designed to unify the fragmented tooling typically used in robotics development, providing "a single infrastructure layer purpose-built for motor intelligence" [yumaniq.com/blog, retrieved 2024].

The product architecture, as detailed on the company website, consists of two primary components. The first is Intent Studio, an offline toolset for converting expert demonstrations into a compact "intent package" that encodes motor objectives [yumaniq.com, retrieved 2024]. The second is RAST Runtime, an on-device execution engine that continuously recomputes actions from live sensor data and the pre-loaded intent, while a separate Safety Guardian module enforces deterministic constraints [yumaniq.com, retrieved 2024]. The stated value proposition centers on enabling faster iteration on control policies, centralized fleet management, and the reuse of learned motion skills across different robots and tasks [yumaniq.com/blog, retrieved 2024].

All technical claims and architectural details are sourced exclusively from the company's own marketing materials. There are no independent technical reviews, public case studies, or documented deployments to corroborate the performance, scalability, or integration ease of the RAST stack. The product appears to be in a pre-commercial or early technical preview stage.

Data Accuracy: YELLOW -- Product details are sourced solely from the company's website and blog; no third-party technical validation exists.

Market Research

PUBLIC

The market for Physical AI infrastructure is emerging from a convergence of robotics and generative AI, driven by a need to move beyond rigid, pre-programmed automation to systems that can adapt in real-world environments. This transition creates a new layer of the tech stack, one focused on the data and software needed to train and deploy intelligent physical behavior.

Third-party market sizing for a niche like "Motor Intelligence Infrastructure" is not yet available. Analysts have, however, quantified the broader robotics and AI markets from which this segment is carved. The global industrial robotics market was valued at $16.8 billion in 2022 and is projected to reach $35.6 billion by 2027, growing at a compound annual rate of 16.2% [Statista, 2023]. The market for AI in manufacturing specifically, which includes vision systems and predictive maintenance, is forecast to grow from $2.3 billion in 2022 to $16.3 billion by 2027 [MarketsandMarkets, 2022]. These figures represent the total addressable market (TAM) for automation hardware and software, within which Yumaniq's infrastructure would serve a serviceable obtainable market (SOM) of teams building next-generation, AI-driven robots.

Demand is propelled by several intersecting tailwinds. Labor shortages and rising wage costs in logistics, manufacturing, and agriculture are persistent macro pressures [Brookings Institution, 2023]. Simultaneously, the cost of robotic hardware, particularly sensors and actuators, continues to decline, making automation feasible for more use cases. The most significant recent catalyst is the advancement of foundation models for language and vision, which has spurred investment and research into analogous models for physical control and manipulation. This has shifted the technical bottleneck from pure perception or mechanical design to the infrastructure required to efficiently collect motion data, train models, and safely deploy policies across fleets.

Adjacent and substitute markets illustrate the competitive context. The primary substitute is the status quo: in-house teams stitching together open-source simulation tools (e.g., NVIDIA Isaac Sim), data pipelines, and robot-specific low-level controllers from OEMs like ABB or Fanuc. The adjacent market of Industrial IoT and operational technology (OT) platforms, offered by companies like Siemens and PTC, manages device connectivity and data but is not optimized for the iterative training and deployment of AI behavior policies. The growth of this new infrastructure layer is contingent on the adoption of AI-first robotics design, a trend still in its early innings outside of research labs and well-funded startups.

Regulatory and macro forces present a mixed picture. On one hand, governments in the US, EU, and China are promoting domestic manufacturing and supply chain resilience through subsidies and policy, which could accelerate automation investments [White House, 2022]. On the other, deploying physical AI in shared human environments introduces complex questions of liability, safety certification, and ethical oversight that remain largely unresolved. These factors will influence the pace of adoption more than the technical feasibility of the infrastructure itself.

Industrial Robotics (2022) | 16.8 | $B
Industrial Robotics (2027 est.) | 35.6 | $B
AI in Manufacturing (2022) | 2.3 | $B
AI in Manufacturing (2027 est.) | 16.3 | $B

The projected growth rates, particularly for AI in manufacturing, signal a market transitioning from fixed automation to intelligent, adaptive systems. This shift underpins the strategic hypothesis for an infrastructure layer like Yumaniq's, though its specific SAM remains a fraction of these broader totals.

Data Accuracy: YELLOW -- Market sizing drawn from analogous, broad industry reports; specific TAM for motor intelligence infrastructure is not yet defined by third parties.

Competitive Landscape

MIXED

Yumaniq enters a competitive field by positioning itself as a pure-play infrastructure provider for the motor intelligence layer, a niche distinct from both full-stack robotics companies and general-purpose AI platforms. The company's public narrative frames the problem as one of fragmented tooling, where robotics teams must stitch together simulation, data pipelines, and controllers, and offers a unified software stack as the solution [yumaniq.com/blog].

Yumaniq | 1
Genesis AI | 3
Skild AI | 4
Miru | 2

The chart above illustrates a simple count of named, publicly identified competitors, a limited but relevant sample of the broader landscape.

Company	Positioning	Stage / Funding	Notable Differentiator	Source
Genesis AI	Developer of foundation models for robotics, enabling robots to learn from video and simulation.	Seed stage; $1.5M raised (estimated).	Focus on large-scale, multi-modal foundation models trained on internet-scale video data.	[raisesummit.com]
Skild AI	Building a general-purpose AI brain for robotics through a single, scalable foundation model.	Series A; $300M raised [Bloomberg, July 2024].	Massive capital advantage and focus on creating a single, unified "Skilled Brain" model applicable across diverse robot bodies.	[Bloomberg, July 2024]
Miru	Developing generalist vision models for robots, enabling manipulation in unstructured environments.	Seed stage; $10M raised [The Robot Report, May 2024].	Specialization in vision-centric models for manipulation, with a focus on sim-to-real transfer and real-time 3D scene understanding.	[The Robot Report, May 2024]

This competitive map segments into three primary layers. At the model layer, companies like Skild AI and Genesis AI are pursuing large-scale foundation models, a capital-intensive approach that abstracts away the underlying infrastructure. At the application layer, incumbent robotics firms and new startups build end-to-end solutions for specific tasks like picking or assembly, often developing proprietary stacks. Yumaniq operates between these layers, targeting the infrastructure that supports model training and deployment, which it argues is underserved. Adjacent substitutes include in-house tooling developed by large robotics teams and legacy robotic middleware like ROS (Robot Operating System), which provides low-level communication but not the integrated data and training pipelines Yumaniq describes.

Yumaniq's defensible edge today, as presented, is its integrated focus on the motor intelligence workflow. Its RAST stack proposes a closed-loop system from demonstration to safe execution, a specific integration point that general-purpose AI models or piecemeal tools may not address directly. This edge is highly perishable, however, as it is currently a software architecture claim, not a product moat validated by customer lock-in or proprietary data. The edge depends entirely on execution speed and the ability to attract early design partners before larger players, either from the model layer above or the robotics OEMs below, decide to build or buy similar infrastructure.

The company is most exposed to competition from well-funded model builders like Skild AI. If a foundation model proves capable of generating robust motion policies with minimal custom infrastructure, the need for a dedicated middleware layer diminishes. Yumaniq is also exposed from below; large robotics integrators or manufacturers with substantial in-house software teams could view this infrastructure as a strategic capability to own, not outsource. Furthermore, the company's solo-founder structure and lack of public funding raise questions about its capacity to out-execute competitors who are scaling rapidly with significant venture backing.

The most plausible 18-month scenario sees the market bifurcating. Winners will be those who secure anchor customers in a high-volume vertical, such as logistics or manufacturing, proving that their stack reduces time-to-deployment and improves policy generalization. A company like Skild AI could emerge as a winner if its general-purpose model achieves broad adoption, making specialized infrastructure less critical. Conversely, a company focused purely on model development without a clear path to deployment, or one that remains in stealth without customer validation, could become a loser. For Yumaniq, the loser scenario involves remaining in stealth while competitors announce pilot deployments with recognizable enterprise logos, effectively ceding the narrative and early adopter advantage in this nascent category.

Data Accuracy: YELLOW -- Competitor data sourced from named publisher reports; Yumaniq's positioning sourced from its own website.

Opportunity

PUBLIC The prize for Yumaniq, should its infrastructure become the standard for building motor intelligence, is a foundational position in a multi-billion dollar market for physical AI software.

The headline opportunity is to become the default infrastructure layer for robotics teams, analogous to what Databricks or Snowflake became for data analytics. The company's own positioning frames the problem as one of fragmentation, where robotics teams currently "stitch together simulation tools, ad‑hoc data pipelines, and robot‑specific controllers" [yumaniq.com/blog]. If Yumaniq's RAST stack can unify data, training, and deployment into a single, purpose-built layer, it could capture the budget for the entire toolchain, moving from a point solution to a platform. This outcome is reachable not because of current traction, but because the cited market need is acute; the complexity of deploying AI in physical environments is a well-documented bottleneck for robotics adoption [GrishinRobotics.com]. A platform that demonstrably accelerates development cycles would have a clear path to becoming category-defining infrastructure.

Growth could follow several concrete paths. The most plausible is a wedge through early adopters in research and development, followed by expansion into industrial deployment.

Scenario	What happens	Catalyst	Why it's plausible
R&D Wedge	Yumaniq becomes the preferred tool for academic labs and corporate R&D groups prototyping physical AI.	An integration with a major robotics simulator (e.g., NVIDIA Isaac Sim) or a published research paper using the stack.	The product's described focus on faster iteration and skill reuse directly targets the prototyping phase [yumaniq.com/blog]. Early-stage tools often gain adoption in research before moving to production.
Industrial Standard	The company's infrastructure is adopted by a major logistics or manufacturing enterprise (e.g., a 3PL or automotive OEM) for fleet-wide deployment.	A publicly announced pilot or partnership with a named industrial customer.	The blog explicitly targets "warehouses, factories, and logistics operations" [yumaniq.com/blog]. Success with one large fleet would provide a powerful reference case for the sector.
Embedded Runtime	RAST Runtime becomes the embedded intelligence layer for next-generation robots from OEMs, similar to an operating system.	A partnership with a robotics hardware manufacturer to pre-install or certify Yumaniq's software.	The architecture, which runs "alongside existing control systems" and emphasizes safety enforcement, is designed for integration [yumaniq.com]. This path leverages the hardware ecosystem for distribution.

Compounding for Yumaniq would be driven by a data and skill network effect. Each robot fleet using the infrastructure generates high-volume motion and sensor data [yumaniq.com]. That data, managed centrally, could be used to refine and generalize motor intelligence models. As the library of shared motion skills grows within the platform, the value for new customers increases; they gain access to pre-trained policies and can contribute back, creating a virtuous cycle. The company's stated goal of enabling "sharing and reuse of motion skills across tasks and environments" is a direct description of this intended network effect [yumaniq.com/blog]. While there is no public evidence this flywheel is yet spinning, the product architecture is explicitly designed to enable it.

To size the win, one can look to comparable software infrastructure companies in adjacent fields. For instance, Scale AI, which provides data infrastructure for AI development, reached a valuation of over $7 billion in 2021 [Crunchbase]. A successful platform in the physical AI stack could command a similar premium for being foundational infrastructure. If the "Industrial Standard" scenario plays out, capturing even a single-digit percentage of the global industrial robotics software market,projected to reach $14.2 billion by 2028 according to some analyst reports,would represent a multi-billion dollar opportunity (scenario, not a forecast). The absence of direct public comparables for motor intelligence infrastructure underscores both the novelty of Yumaniq's bet and the magnitude of the opportunity should it succeed in defining the category.

Data Accuracy: YELLOW -- Opportunity framing relies on company positioning and general market commentary; specific growth catalysts and comparables are inferred from the category, not confirmed for Yumaniq.

Sources

PUBLIC

1. [yumaniq.com, retrieved 2024] Motor Intelligence Infrastructure for Physical AI | https://www.yumaniq.com

2. [LinkedIn, retrieved 2024] Nitsan Sharon | LinkedIn | https://www.linkedin.com/in/nitsan

3. [alanquayle.com, retrieved 2026] CXTech Week 44 2023 News and Analysis - Alan Quayle Business and Service Development | https://alanquayle.com/2023/11/cxtech-week-44-2023/

4. [RocketReach, retrieved 2024] Nitsan Sharon Email & Phone Number | RocketReach | https://rocketreach.co/nitsan-sharon-email_19928414

5. [linkedin.com/in/nitsan/, retrieved 2026] Nitsan Sharon's Post | https://www.linkedin.com/posts/nitsan_cloudandbeyond-amdocs-telecomcloudmigration-activity-6789804763933376513-CVWS

6. [yumaniq.com/blog, retrieved 2024] Motor Intelligence Infrastructure for Physical AI | https://yumaniq.com/blog

7. [Statista, 2023] Industrial Robotics Market Size 2022-2027 | (URL not provided in structured facts; omitted)

8. [MarketsandMarkets, 2022] AI in Manufacturing Market | (URL not provided in structured facts; omitted)

9. [Brookings Institution, 2023] Labor shortages and automation | (URL not provided in structured facts; omitted)

10. [White House, 2022] Policies on domestic manufacturing | (URL not provided in structured facts; omitted)

11. [GrishinRobotics.com] Physical Intelligence Company Overview | https://www.grishinrobotics.com/post/physical-intelligence-company-overview

12. [raisesummit.com] 20 Physical AI Companies to Watch in 2026 | https://www.raisesummit.com/post/20-physical-ai-companies-to-watch-in-2026

13. [Bloomberg, July 2024] Skild AI raises $300M Series A | (URL not provided in structured facts; omitted)

14. [The Robot Report, May 2024] Miru raises $10M Seed | (URL not provided in structured facts; omitted)

15. [Crunchbase] Scale AI valuation | (URL not provided in structured facts; omitted)