Skip to content
HOMETECH DIRECTORYNEWS & DEVELOPMENTSRELOCATEEMPLOYERSNEIGHBORHOODSABOUTSchedule a Call

Invisible AI

Invisible AI runs computer vision on factory floors. South San Francisco HQ, Ann Arbor office at 330 East Liberty, and a global automotive OEM deployment.

A2Hub Editorial· Rusty Roof Media· Updated May 4, 2026

Invisible AI

Invisible AI is the industrial computer vision company that figured out how to watch a factory floor without putting its data in somebody else's cloud. The company builds edge camera systems that run deep learning locally on the device, mount above assembly stations, and produce live analytics on human motion, cycle time, ergonomics, and process deviation. The corporate headquarters sits in South San Francisco, but the product was always going to require a real Midwest presence, and the Ann Arbor office at 330 East Liberty Street is the operational hub for that thesis. The company's largest publicly referenced deployment is a global automotive OEM running more than 1,500 Invisible AI edge devices across 14 North American plants. That scale is unusual for industrial computer vision at this stage and is the proof point the company's pitch rests on. For Ann Arbor, Invisible AI is a clean example of coastal capital underwriting a real in-town engineering presence because the customers live here.

Quick facts

  • Founded: 2018
  • Founders: Eric Danziger (CEO), Prateek Sachdeva (CTO)
  • HQ: South San Francisco, California
  • Ann Arbor office: 330 East Liberty Street, downtown Ann Arbor
  • Headcount: Multi-dozen across the San Francisco and Ann Arbor offices; specific breakdowns not publicly disclosed
  • Funding: Approximately $30M+ across seed and Series A disclosures
  • Key investors: 8VC, M12 (Microsoft's venture fund), Sierra Ventures
  • Current leader: Eric Danziger, CEO and co-founder
  • Sector tags: AI, computer vision, manufacturing, industrial AI, edge AI
  • Careers: invisible.ai/careers
  • Website: invisible.ai
  • LinkedIn: linkedin.com/company/invisible-ai

What Invisible AI makes

Invisible AI ships a vertically integrated edge computer vision platform built for industrial environments. Hardware, models, and software are sold together as a single deployable unit rather than as a cloud service that requires the customer to pipe video offsite.

Edge devices. Purpose-built camera and compute units that mount above assembly stations, workcells, or specific task areas on a plant floor. The device runs deep learning inference locally, which is what makes the system viable inside plants where bandwidth, IT policy, and data sovereignty all push back against streaming video to an external cloud. Most major auto OEMs prohibit continuous video egress from plant networks, so the edge-first architecture is not a technical preference. It is a commercial requirement.

Motion analytics for human work. The flagship application analyzes how human operators perform tasks. Cycle times, deviation from the documented standard work, ergonomic risk indicators, missed or out-of-order steps, and line-balancing signals. Plant engineers use the data to rebalance stations, design training interventions, and reduce the kind of repetitive-motion injuries that drive workers' compensation exposure.

Quality and assurance applications. Beyond the motion focus, the platform covers quality inspection, presence and absence detection, fastener and torque validation, and process verification. That widens the product into general-purpose plant computer vision rather than a pure ergonomics tool.

Software platform. A web application and APIs that aggregate data across devices, plants, and lines into the dashboards plant managers and process engineers actually use. The company positions this as the "always-on industrial engineer," a framing that travels well with the operations-research heritage of the auto manufacturing industry.

Customer concentration tilts heavily toward automotive and automotive tier-one suppliers, with expansion into general manufacturing, food and beverage, and distribution. Pricing is per-device plus software subscription, which aligns the company's revenue with deployment breadth rather than one-off integrator engagements.

The Ann Arbor story

Invisible AI was founded in 2018 by Eric Danziger and Prateek Sachdeva. Both came out of computer vision and machine learning backgrounds with prior experience at Clarifai, and the founding thesis was sharp. Most computer vision being deployed in industrial settings was either expensive systems-integrator work or science experiments that never got to scale, and there was room for a productized hardware-plus-software platform that a plant could install in days and start getting useful signals from in weeks.

The company chose South San Francisco for the corporate headquarters, for the usual reasons, coastal capital, AI research network, hardware supply chain in the Bay Area. But the customer map pointed the other direction. The auto OEMs and tier-ones where industrial computer vision has the most compelling near-term ROI are concentrated in southeast Michigan, Ohio, and the broader Midwest manufacturing belt. Putting the Ann Arbor office at 330 East Liberty Street is a deliberate move. The address sits inside the downtown core a few blocks from Cisco's Duo team, Censys, and Nutshell, walking distance to Main Street, and under an hour's drive from most major Detroit-area plants.

The Ann Arbor office runs a mix of customer engineering, customer success, and engineering staff focused on the manufacturing customer base. The U-M College of Engineering and the Robotics Institute supply the talent pipeline. Ann Arbor SPARK and the broader manufacturing-tech ecosystem, including Toyota Motor North America R&D, Mcity, KLA, and the American Center for Mobility, provide the network density. The office is the kind of in-town satellite that justifies itself on customer proximity and engineering recruiting rather than on headquarters politics.

The broader pattern matters. Invisible AI is one of a small group of coastal-headquartered AI startups that have chosen to maintain meaningful Ann Arbor footprints because the customer base or talent base sits in southeast Michigan. That includes Sight Machine and Deepgram, and it is one of the more durable arguments for Ann Arbor's strategic relevance in the AI cycle. The city sits inside one of the world's largest industrial manufacturing footprints, and applied AI on top of that footprint cannot be run from a coastal desk alone.

Funding and exits

Invisible AI has raised approximately $30M+ across seed and Series A rounds. 8VC led the early investment. Microsoft's M12 venture fund and Sierra Ventures have participated alongside. The company has not publicly disclosed revenue, detailed headcount, or valuation since the Series A close. Any subsequent extension or Series B activity should be verified against current filings before publication.

There has been no exit. The company remains private and growth-stage. Strategic exit candidates, if one were to plot them, would include industrial automation incumbents (Rockwell, Siemens, ABB), automotive software providers, and the hyperscaler industrial AI offerings, though none of those paths is under public discussion. Management posture in interviews reads as committed to building a standalone business rather than optimizing for acquisition.

Culture and hiring

Invisible AI hires computer vision engineers, embedded systems engineers, customer engineering specialists, field operations staff, and customer success managers. The work profile is unusual for a computer vision company because it sits at the intersection of ML modeling, embedded systems, and on-site industrial operations. A candidate who can hold both a PyTorch notebook and a hard hat is exactly the profile.

Work model is hybrid with hub concentrations in South San Francisco and Ann Arbor. Ann Arbor roles skew toward customer engineering and field ML, which match the office's function as the operational hub for the Midwest manufacturing customer base. Compensation is competitive with Series A industrial AI companies; equity is private-company stock with standard vesting. Public Glassdoor and Blind signal is limited in volume and trends neutral to positive on technical work and mission clarity.

Why they matter for Ann Arbor tech

Invisible AI is the case study for why the Ann Arbor thesis in industrial AI has legs. The city is within an hour of billions of dollars of live manufacturing output, U-M feeds engineering talent into the pipeline, and downtown is walkable for the kind of customer-engineering rhythm that plant rollouts require. For the Ann Arbor AI pillar, Invisible AI sits alongside Voxel51 as the two most commercially mature computer vision companies with real Ann Arbor engineering presence. Voxel51 owns the infrastructure and data layer, Invisible AI owns a specific industrial application layer, and together they make the case that Ann Arbor can be a first-class location for computer vision as an industry segment.

FAQ

What does Invisible AI do? Invisible AI builds computer vision systems that monitor human motion and industrial processes on factory floors. The platform uses edge devices installed at assembly stations, runs deep learning inference locally on the device, and produces analytics on cycle time, ergonomic risk, missed steps, and line balance for plant operators.

Where is Invisible AI headquartered? Invisible AI is headquartered in South San Francisco, California. The company also operates an Ann Arbor office at 330 East Liberty Street that serves its Midwest manufacturing customer base.

Who founded Invisible AI? Invisible AI was founded in 2018 by Eric Danziger (CEO) and Prateek Sachdeva (CTO), both with backgrounds in computer vision and machine learning, including prior time at Clarifai.

Who are Invisible AI's customers? The company's largest publicly referenced customer is a global automotive OEM running more than 1,500 Invisible AI edge devices across 14 North American plants. Other customers span automotive tier-ones, general manufacturing, and adjacent industrial verticals.

Why is Invisible AI in Ann Arbor? The Ann Arbor office is the operational hub for the company's automotive and Midwest manufacturing customer base. Most major southeast Michigan auto plants are within an hour's drive, and U-M supplies the engineering and robotics talent pipeline. The office is at 330 East Liberty Street in downtown Ann Arbor.

Who competes with Invisible AI? On the industrial computer vision side, Invisible AI competes with Drishti (acquired by R-Go Robotics), Cognex's deep learning line, Landing AI, Retrocausal, and larger industrial automation providers including Rockwell and Siemens that have bolted computer vision into existing MES stacks.

Related Ann Arbor AI and industrial tech companies

  • Voxel51 is the Ann Arbor headquartered computer vision dataset and MLOps platform at the data and model-lifecycle layer.
  • Sight Machine is the manufacturing analytics and AI platform with offices in San Francisco and Ann Arbor, running an adjacent industrial AI thesis.
  • MemryX is the Ann Arbor edge AI accelerator chip company; a natural hardware partner for industrial CV at the device layer.
  • Toyota Motor North America R&D anchors the 50-year automotive engineering footprint in Ann Arbor that underpins the local industrial AI customer base.
  • The Ann Arbor AI companies hub covers the full regional roster. The site homepage is the master directory.