The most expensive sensor in a factory is the one you have to buy. Khenda, a Turkish startup founded in 2021, is betting that the cheapest one is the camera you already have. Their platform uses computer vision to turn ordinary video from smartphones or security cameras into a stream of structured data: cycle times, work instructions, and line-balancing analyses. It is a pragmatic, capital-light wedge into manufacturing efficiency. The more ambitious bet, however, is what happens to all that video data next. Khenda is also positioning itself as the data infrastructure layer for Physical AI, converting industrial video into the structured training data needed to teach humanoid robots how to work [khendarobotics.com, Unknown].
A pragmatic wedge into the factory floor
Khenda's initial product is a continuous improvement platform built for manufacturing engineers [Perplexity Sonar Pro Brief, Unknown]. The value proposition is straightforward: avoid expensive hardware retrofits or PLC integrations. Instead, upload video of a production line, and the AI generates time studies, identifies non-value-added steps, and suggests optimized standard operating procedures. For one client, ISAIC, the company claims its analysis saved 80% of the time compared to manual methods and improved line efficiency by 15% [LinkedIn, Aykan Ekici, Retrieved 2026]. The software modules read like a lean manufacturing checklist: Digital Time Study, MTM analysis, a Lean Assistant, and a Vision MES for real-time production monitoring [Perplexity Sonar Pro Brief, Unknown]. The target buyer is clear: an operations or industrial engineering team under pressure to increase throughput without a capital expenditure request.
The team and the transatlantic bridge
The company is led by co-founders and co-CEOs Çağkan and Aykan Ekici. Public profiles highlight Aykan Ekici's over eight years of experience in computer vision projects [Dealroom.co, Unknown]. While the founders' prior exit track record isn't detailed in the public record, their strategic geography is noteworthy. Headquartered in Istanbul, Khenda has established a virtual desk in Ann Arbor, Michigan, with support from the economic development organization Ann Arbor SPARK [Purpose Jobs, Retrieved 2026]. This bridgehead into the American Midwest, a heartland of advanced manufacturing, signals a deliberate push beyond its initial Turkish market focus. The move is less about a physical office and more about accessing the strategic networks and potential pilot customers in a key industrial region.
Funding a dual-track roadmap
In 2024, Khenda closed a seed funding round of $1.75 million [Khenda Blog, 2024]. The round was led by Diffusion Capital Partners and included Simya VC and the Koç Group Companies' First Private Equity Investment Fund [Khenda Blog, 2024]. This capital is fueling the company's two-track strategy. The first, and currently revenue-generating, track is the continuous improvement SaaS platform. The second, more future-oriented track is the development of a structured data pipeline for robotics training. The company's website frames this as building "the data layer for Physical AI," explicitly aiming to convert real-world industrial video into training datasets for humanoid robots [khendarobotics.com, Unknown]. The seed funding suggests investors are backing both the near-term manufacturing tool and the longer-term infrastructure bet.
Seed Round (2024) | 1.75 | M USD
Where the wheels could come off
For any early-stage company pursuing a dual-track strategy, the primary risk is focus. The manufacturing efficiency software market is already crowded with established players and well-funded startups. Khenda's wedge of using standard video is clever, but it must prove it can scale sales and achieve durable retention in a competitive space before the runway from its seed round expires. The robotics data infrastructure angle, while strategically interesting, is a longer-term play that depends on the broader adoption of humanoid robots in industrial settings, a trend still in its early innings. The company's public traction is currently demonstrated through claimed efficiency gains for a single client, ISAIC; the path to a diversified, referenceable customer base is the critical next proof point.
- Market focus. The manufacturing software space includes deep-pocketed incumbents and specialized AI vision startups. Khenda must clearly define and dominate a specific vertical or use case.
- Data pipeline maturity. The value of robotics training data is contingent on volume, annotation quality, and format standardization. Building this as a scalable product is a separate engineering challenge from the core video analytics platform.
- Capital efficiency. With $1.75 million in seed funding (estimated) [CompanyCheck, Dec 2024], the company must carefully allocate resources between its immediate SaaS business and its future-facing data infrastructure build.
The next twelve months
The immediate milestones for Khenda are commercial and product-led. Success will be measured by landing and publicly naming manufacturing customers in its new target market of North America, moving beyond the single cited case study. On the product side, watch for a clearer articulation of how the manufacturing efficiency platform feeds the robotics data pipeline, potentially through a dedicated developer API or a new product SKU. Another seed extension or a Series A round within the next 12-18 months is likely, contingent on demonstrating growth in annual recurring revenue from its SaaS platform. The Ann Arbor bridgehead will be tested for its ability to generate qualified leads and strategic partnerships.
Khenda's ideal customer profile today is a mid-sized manufacturing plant with high labor costs and capacity constraints, where the industrial engineering team is tasked with driving efficiency but lacks the budget for sensor-based automation. The realistic competitive set splits into two lanes. For its aspirational data infrastructure layer, the competition is less direct but includes any company aggregating real-world robotic demonstration data, such as Covariant or the internal data operations of large robotics labs. Khenda's bet is that owning the primary video data source on the factory floor gives it a unique dataset advantage for both fights.
Sources
- [khendarobotics.com, Unknown] Khenda | The Data Infrastructure for Physical AI | https://khendarobotics.com/
- [Perplexity Sonar Pro Brief, Unknown] Khenda product and market overview
- [LinkedIn, Aykan Ekici, Retrieved 2026] Client case study for ISAIC
- [Dealroom.co, Unknown] Khenda company information and team background
- [Purpose Jobs, Retrieved 2026] Why This Turkish AI Startup Wanted in on Ann Arbor’s Tech Scene | https://www.purpose.jobs/blog/turkish-ai-startup-moves-to-ann-arbor
- [Khenda Blog, 2024] Khenda seed funding announcement
- [CompanyCheck, Dec 2024] Khenda funding and corporate profile
- [Structured Facts] Competitor list