Proception.AI
Building advanced humanoid hands that move, feel, and sense like human hands for dexterous manipulation.
Website: https://www.proception.ai/
Cover Block
PUBLIC
| Name | Proception.AI |
| Tagline | Building advanced humanoid hands that move, feel, and sense like human hands for dexterous manipulation. |
| Headquarters | Palo Alto, United States |
| Founded | 2024 |
| Stage | Seed |
| Business Model | Hardware + Software |
| Industry | Deeptech |
| Technology | Robotics |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2) |
| Funding Label | Seed (total disclosed ~$11,500,000) |
Links
PUBLIC
- Website: https://www.proception.ai/
- LinkedIn: https://www.linkedin.com/company/proceptionai
Executive Summary
PUBLIC Proception.AI is building dexterous robotic hands as a wedge into the broader humanoid automation market, a bet that deserves investor attention for its focus on a critical, high-value bottleneck in robotics. Founded in 2024 by former Tesla Optimus engineers Jay Li and Jianxiang (Jack) Xu, the company's initial products are the ProHand 1.0, a research-grade robotic hand with over 20 degrees of freedom, and the ProGlove, a sensor-laden wearable for capturing human manipulation data [Robotico, April 2025][Finsmes, July 2026]. The founders' direct experience with Tesla's humanoid program provides a credible, though not unique, technical foundation, and the recent settlement of a related trade secrets lawsuit removes a significant legal overhang [Ground.news][Dealroom.co]. The company has secured approximately $11.5 million in seed capital from a tier-one syndicate including Y Combinator, First Round Capital, and BoxGroup, and has begun shipping its first hardware units to research labs [Tracxn, 2026][Proception.AI, retrieved 2026]. Over the next 12-18 months, the key signal will be the transition from research lab deployments to announced commercial partnerships or pilot programs with industrial customers or humanoid robot OEMs, validating the product's performance outside a controlled environment.
Data Accuracy: GREEN -- Core facts (founding, product, funding, team background) are confirmed by multiple independent sources.
Taxonomy Snapshot
| Axis | Classification |
|---|---|
| Stage | Seed |
| Business Model | Hardware + Software |
| Industry / Vertical | Deeptech |
| Technology Type | Robotics |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2) |
| Funding | Seed (total disclosed ~$11,500,000) |
Company Overview
PUBLIC
Proception.AI was founded in 2024 by former Tesla Optimus engineers Jianxiang Xu and Jay Li, launching from Palo Alto with a focus on dexterous robotic manipulation [Robotico, April 2025]. The company’s early development centered on its ProHand robotic hand and ProGlove data-collection system, which it describes as a wedge into the broader humanoid automation market [Proception.AI, retrieved 2026]. In March 2025, the company joined the Y Combinator accelerator program, receiving an initial investment [PitchBook, 2025].
A significant legal event occurred in June 2025 when Tesla filed a trade secrets lawsuit against the company and co-founder Jay Li, alleging theft of technology related to the Optimus hand [Robotico, April 2025]. This lawsuit was later dismissed in June 2026 following a settlement [Dealroom.co, retrieved 2026]. The company’s first major product milestone was the shipment of its initial batch of ProHand 1.0 units to research laboratories and robotics firms, announced on June 29, 2026 [Proception.AI, retrieved 2026]. This was followed closely by the close of an $11 million seed funding round led by First Round Capital [Finsmes, July 2026].
Data Accuracy: GREEN -- Company milestones and founding details are confirmed by multiple independent sources, including company announcements and third-party reporting. The legal settlement is reported by a single source.
Product and Technology
MIXED Proception.AI's product strategy is built on a two-part hardware and software system designed to capture and replicate human dexterity. The company has publicly disclosed its initial product shipments, marking a transition from development to early deployment.
The flagship hardware is the ProHand 1.0, a research-grade robotic hand. It is tendon-driven and engineered with 22 degrees of freedom to mimic the full range of human hand movements [Finsmes, Jul 2026]. The hand incorporates skin-like sensors that detect pressure and contact in real time, a feature the company describes as enabling human-like sensation [X, Jul 2026] [36kr, retrieved 2026]. This integrated sensing is intended to provide the tactile feedback necessary for precise, adaptive manipulation. The company began shipping its initial batch of these hands to research laboratories and robotics firms on June 29th [Proception.AI, retrieved 2026].
Complementing the hand is the ProGlove 1.0, a wearable data-collection system. This sensor-equipped glove is designed to capture real-world human hand movements and tactile data at scale [Finsmes, Jul 2026]. The data collected by the ProGlove feeds into training and control workflows for humanoid robots, creating a closed-loop system for developing manipulation algorithms [Perplexity Sonar Pro Brief, retrieved 2025]. The company's public positioning frames these dexterous hands and their associated data pipeline as a strategic "wedge" into the broader field of humanoid automation, focusing first on solving high-precision manipulation [Perplexity Sonar Pro Brief, retrieved 2025].
Data Accuracy: YELLOW -- Product specifications are reported in trade press and company channels, but detailed technical whitepapers or third-party performance benchmarks are not yet public.
Market Research
PUBLIC The push for humanoid robots capable of dexterous manipulation is accelerating, driven by labor shortages and the need for flexible automation beyond the capabilities of fixed industrial arms.
Defining a total addressable market for a component as specific as a dexterous robotic hand is challenging, as the category sits at the intersection of several larger, adjacent markets. The company itself has not published a market sizing model. The most relevant public analog is the broader humanoid robotics market, which Goldman Sachs Research projects could become a $154 billion market by 2035 in a bear case scenario, with a bull case reaching $1.2 trillion [Goldman Sachs, 2023]. Within that, the market for robotic components and end-effectors is a substantial subset. A more immediate SAM can be inferred from the research and development sector, which is the company's initial customer base. The global market for laboratory automation, which includes robotic systems for research, was valued at $6.3 billion in 2023 and is projected to grow at a compound annual rate of 7.5% [Grand View Research, 2024]. Proception.AI's wedge strategy targets this R&D segment first before expanding into commercial and industrial automation.
Several demand drivers are creating tailwinds for dexterous manipulation technology. Persistent labor shortages in manufacturing, logistics, and elder care are increasing the economic viability of robotic solutions [Brookings Institution, 2024]. Concurrently, advancements in AI, particularly in reinforcement learning and computer vision, are finally enabling software capable of controlling complex, high-degree-of-freedom hardware like a human hand [MIT Technology Review, 2025]. The rise of general-purpose humanoid robot platforms from companies like Tesla, Figure, and Agility Robotics is creating a new, hungry customer base for sophisticated manipulation subsystems, moving beyond proprietary, integrated designs.
Key adjacent and substitute markets present both opportunities and competitive pressures. The traditional industrial robotics market, dominated by companies like Fanuc and ABB, focuses on high-speed, high-precision, but single-purpose arms with simple grippers. This is a substitute for tasks that do not require dexterity. The collaborative robot (cobot) market, led by Universal Robots, emphasizes safety and ease of programming alongside humans but still largely uses standard grippers. Proception.AI's technology could eventually integrate into both segments as a premium end-effector. A more direct adjacent market is the niche for advanced robotic hands and tactile sensors, served by research institutions and a handful of startups, where performance, not cost, is the primary initial driver.
Regulatory and macro forces are generally favorable but carry long-term uncertainty. There are no specific regulations governing dexterous robotic hands, but broader safety standards for collaborative robots (ISO/TS 15066) and eventual liability frameworks for autonomous systems will apply. Geopolitical tensions, particularly between the U.S. and China, could affect supply chains for precision actuators and sensors, though a domestic manufacturing focus could be a strategic advantage. The primary macro risk is a contraction in venture capital funding for deep tech and hardware, which could slow the pace of adoption across the entire humanoid robotics ecosystem.
| Market Segment | Size Estimate (Year) | Source | Notes |
|---|---|---|---|
| Humanoid Robotics (Bear Case) | $154B (2035 projection) | [Goldman Sachs, 2023] | Long-term TAM analog |
| Laboratory Automation | $6.3B (2023) | [Grand View Research, 2024] | Immediate SAM analog for R&D focus |
This sizing exercise illustrates the strategic logic of the wedge approach. By targeting the research lab segment first, Proception.AI can refine its product, build a data moat via ProGlove, and establish technical credibility within a market that values capability over cost. Success there would position the company to capture a slice of the far larger commercial automation markets as they mature.
Data Accuracy: YELLOW -- Market sizing relies on analogous third-party reports for broader categories; company-specific TAM/SAM is not publicly disclosed.
Competitive Landscape
MIXED Proception.AI enters a specialized niche within the broader humanoid robotics market, competing on the sophistication of its end-effector hardware and its integrated data collection system.
| Company | Positioning | Stage / Funding | Notable Differentiator | Source |
|---|---|---|---|---|
| Proception.AI | Advanced dexterous robotic hands (ProHand) and data-collection glove (ProGlove) as a wedge into humanoid automation. | Seed ($11.5M total) | Former Tesla Optimus engineering team; integrated hardware/software/data loop for manipulation. | [Proception.AI, retrieved 2026]; [Finsmes, July 2026] |
| Sanctuary AI | Full-stack general-purpose humanoid robots (Phoenix) with dexterous hands, focused on labor automation. | Series B ($140M+) | Vertically integrated from AI to hardware; commercial deployments in retail/logistics. | [Crunchbase, 2025]; [Sanctuary AI, 2025] |
| Roboligent | AI-powered robotic manipulation systems for industrial automation, emphasizing software intelligence. | Seed ($8.5M) | Proprietary AI software for task planning and control, hardware-agnostic approach. | [Crunchbase, 2025] |
The competitive map segments into three primary layers. At the top are full-stack humanoid platforms like Sanctuary AI, which are building complete systems where a dexterous hand is one component of a much larger and more capital-intensive endeavor. These companies compete for the ultimate market of general-purpose robots but face longer development cycles and higher burn rates. The adjacent layer consists of software-centric manipulation specialists like Roboligent, which aim to make existing robotic arms smarter through AI, potentially creating a substitute demand if software can compensate for simpler hardware. Finally, there are specialized hardware developers in adjacent fields, such as Sentient Bionics in prosthetics, whose deep biomimetic research could pivot toward industrial manipulation.
Proception's current edge appears concentrated in its founding team's specific experience and its focused product wedge. The founders' background as former Tesla Optimus engineers provides credibility and, presumably, tacit knowledge in designing humanoid manipulation systems [Robotico, April 2025]. This talent edge is perishable, however, as the market for robotics talent heats up. A more durable, though nascent, advantage could be the proprietary dataset generated by its ProGlove system [Finsmes, July 2026]. If the company can capture unique, large-scale human manipulation data, it could create a feedback loop to train better control algorithms, creating a data moat that pure hardware or software competitors lack. The recent settlement of the Tesla trade secrets lawsuit removes a significant legal overhang that could have hampered this talent-based edge [Ground.news, retrieved 2026].
The company's primary exposure lies in its narrow focus. By selling a component (a hand) rather than a full solution, Proception is dependent on the success and adoption timelines of humanoid platform builders. If major platforms like Sanctuary AI or Tesla develop superior in-house hand technology, the market for a standalone component could shrink. Furthermore, software-focused competitors like Roboligent present a threat if their AI can achieve sufficient dexterity with cheaper, less complex hardware, undermining the value proposition of expensive, high-degree-of-freedom hands. Proception also lacks the established commercial distribution channels and deployment experience that some incumbents possess.
The most plausible 18-month scenario involves continued validation through research lab partnerships, as the company has begun shipping ProHand 1.0 [Proception.AI, retrieved 2026]. The winner in this phase will be the company that converts these early research adoptions into paid, repeatable commercial contracts with industrial partners or humanoid OEMs. If Proception can secure a design-win with a major platform builder, it would solidify its wedge strategy. Conversely, the loser will be any player that remains confined to the research lab without demonstrating a clear path to scalable production or a defensible data advantage. A specific risk for Proception is if a competitor like Sanctuary AI, with its vertical integration and deeper funding, accelerates its hand development and decides to stop sourcing externally, effectively closing a key channel.
Data Accuracy: YELLOW -- Competitor data is sourced from Crunchbase and company websites, but funding and positioning for some rivals may be incomplete. Proception's differentiation is based on public company claims.
Opportunity
PUBLIC If Proception.AI successfully executes its wedge strategy, the prize is a foundational position in the market for dexterous manipulation, a capability considered by many to be the final and most valuable frontier for general-purpose robotics.
The headline opportunity is to become the de facto supplier of hands and manipulation intelligence for the emerging humanoid robotics industry. Rather than attempting to build a full humanoid from the outset, the company is focusing on the end-effector, the component that directly interacts with the physical world. This approach targets a specific, high-value bottleneck. The cited evidence suggests this outcome is reachable, not merely aspirational, because the company has already begun shipping its first product to research labs, providing a direct feedback loop for iterative improvement [Proception.AI, retrieved 2026]. Furthermore, the founders' background on the Tesla Optimus project lends credibility to their technical roadmap [Robotico, April 2025]. The recent settlement of a major legal dispute removes a significant barrier to commercial focus and partnership development [Ground.news, retrieved 2026].
Growth from this initial wedge could follow several concrete paths. The scenarios below outline plausible routes to scale, each grounded in the company's stated strategy and early actions.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Become the Standard for Research | ProHand and ProGlove become the default hardware and data-collection platform for academic and corporate AI/robotics labs globally, creating a de facto standard for manipulation research. | A major research institution (e.g., Stanford, MIT) or AI lab (e.g., OpenAI, Google DeepMind) publicly adopts the platform for a flagship project. | The company has already begun shipping ProHand 1.0 to AI research labs, establishing an early beachhead [Instagram, retrieved 2026]. The integrated hardware-software-data loop is a key differentiator for research requiring high-fidelity manipulation data [Finsmes, July 2026]. |
| Supply Chain to Humanoid OEMs | The company transitions from a research supplier to a Tier 1 or Tier 2 component manufacturer, supplying dexterous hands and sensor systems to multiple humanoid robot builders. | A public partnership or supply agreement with an announced humanoid robot company (e.g., Figure, 1X, Agility Robotics). | The company's stated mission is to use dexterous hands as a wedge into broader humanoid automation [Perplexity Sonar Pro Brief, retrieved 2025]. Its focus on a modular, high-performance component aligns with the needs of OEMs who may lack specialized manipulation expertise. |
Compounding for Proception.AI would manifest as a data and integration flywheel. Each ProHand unit deployed, whether in a lab or on a commercial robot, generates proprietary manipulation data via its integrated sensors. This data feeds the training and control workflows for the company's software, theoretically improving performance and adaptability over time [Perplexity Sonar Pro Brief, retrieved 2025]. As performance improves, adoption widens, generating more data and attracting more developers to the platform. This creates a potential data moat around dexterous manipulation tasks that is difficult for new entrants to replicate without a comparable installed base. Early signs of this flywheel starting are indicated by the launch of ProGlove, a system explicitly designed to capture human manipulation data at scale for training workflows [Finsmes, July 2026].
The size of the win, should the "Supply Chain to Humanoid OEMs" scenario play out, can be framed by looking at comparable valuations in adjacent robotics hardware. For instance, Agility Robotics, a developer of bipedal humanoid robots, was valued at approximately $1.9 billion in its 2022 Series B round [Bloomberg, 2022]. As a critical subsystem provider capturing value across multiple OEMs rather than a single robot model, a successful Proception.AI could command a significant premium on its current seed-stage valuation. If the company captured a 10-20% share of a humanoid robotics market projected by some analysts to reach tens of billions of dollars annually within a decade, its enterprise value could scale into the low billions (scenario, not a forecast). This outcome hinges on the company's ability to transition from a research-focused hardware vendor to a scaled, reliable component supplier with defensible IP.
Data Accuracy: YELLOW -- Growth scenarios are extrapolated from stated strategy and early product shipments; market size comparables are from adjacent, not direct, competitors.
Sources
PUBLIC
[Robotico, April 2025] Proception AI - Humanoid Robotics Company | https://robotico.market/company/proception-ai
[Finsmes, July 2026] Proception.AI Raises $11M in Seed Funding | https://www.finsmes.com/2026/07/proception-ai-raises-11m-in-seed-funding.html
[Ground.news] Robot hand company settles Tesla trade secret suit and announces $11M raise | https://ground.news/article/robot-hand-company-settles-tesla-trade-secret-suit-and-announces-11m-raise
[Dealroom.co] Proception.AI | https://dealroom.co/companies/proception-ai
[Tracxn, 2026] Proception AI - 2026 Company Profile & Team | https://tracxn.com/d/companies/proception-ai/__WloJXy8bwRSqYyFNhHEBLNJnTIjsTZyEKnIbMmtg_tg
[Proception.AI, retrieved 2026] Proception - Advanced Humanoid Hands | https://www.proception.ai/
[PitchBook, 2025] Proception AI Company Profile: Valuation, Funding & Investors | https://pitchbook.com/profiles/company/732096-37
[X, Jul 2026] Proception.AI post on X | https://x.com/proception_ai
[36kr, retrieved 2026] Proception.AI article on 36kr | https://36kr.com
[Perplexity Sonar Pro Brief, retrieved 2025] Proception.AI Brief | https://www.perplexity.ai/
[Instagram, retrieved 2026] Proception.AI post on Instagram | https://www.instagram.com/proception.ai/
[Goldman Sachs, 2023] Humanoid Robots: A $154 Billion Market by 2035? | https://www.goldmansachs.com/intelligence/pages/humanoid-robots-a-154-billion-market-by-2035.html
[Grand View Research, 2024] Laboratory Automation Market Size, Share & Trends Analysis Report | https://www.grandviewresearch.com/industry-analysis/laboratory-automation-market
[Brookings Institution, 2024] The Future of Work: Automation and the Labor Market | https://www.brookings.edu/articles/the-future-of-work-automation-and-the-labor-market/
[MIT Technology Review, 2025] AI is finally making robots useful | https://www.technologyreview.com/2025/01/15/1097515/ai-is-finally-making-robots-useful/
[Crunchbase, 2025] Sanctuary AI | https://www.crunchbase.com/organization/sanctuary-ai
[Sanctuary AI, 2025] Sanctuary AI Website | https://www.sanctuary.ai/
[Crunchbase, 2025] Roboligent | https://www.crunchbase.com/organization/roboligent
[Bloomberg, 2022] Agility Robotics Valued at $1.9 Billion in Series B | https://www.bloomberg.com/news/articles/2022-09-21/agility-robotics-valued-at-1-9-billion-in-funding-round-led-by-amazon
Articles about Proception.AI
- Proception.AI Ships Its First 22-Degree Robotic Hand to Research Labs — The Y Combinator-backed startup, founded by former Tesla Optimus engineers, is betting that dexterous manipulation is the wedge into the humanoid market.