CountAI’s Knit-i brings real-time AI-powered quality inspection directly onto circular knitting machines
CountAI, an Indian industrial AI company building computer-vision systems for textile manufacturing, has won the Intel 2026 Excellence & Innovation Award in the Edge AI Software category for its flagship product, Knit-i.
The award was presented on 13 August at Intel India’s Industry Solution Builders Connect in Bengaluru, an event attended by around 900 manufacturers, technology companies and industry leaders.
CountAI was also among the companies invited to demonstrate its technology at the event and participated in the panel discussion “The AI Ecosystem Playbook: How ISVs, OEMs and Intel Scale Together.”
For CountAI, the recognition is significant for a reason beyond the technology itself: Knit-i is designed to operate directly on the textile production floor, continuously inspecting fabric as it is being knitted rather than detecting defects after the fabric has already been produced.
From finding defects to preventing them
In conventional textile manufacturing, fabric quality is often assessed after it has been knitted and wound into a roll. By the time a defect is discovered, the machine may have continued producing defective fabric for a significant distance.
Knit-i changes the timing of inspection.
An industrial camera mounted on the circular knitting machine continuously observes the fabric as it is formed. An edge-AI computing system analyses the images in real time and identifies defects such as needle lines, holes, elastane breaks and yarn-related variations.
When a critical defect is detected, the system can alert the operator and stop the machine.
The difference is simple but economically important: instead of discovering a problem after the fabric has been made, the mill can act while the problem is still being created.
“A mill owner is not evaluating our algorithm. He is deciding whether he can take inspectors off the table on our word. If our box fails quietly, he is the one who pays for it — not us. Everything we have built is aimed at making sure that never happens.” — Harsha Vardhan, Founder & CEO, CountAI
AI that has to survive the factory floor
For CountAI, industrial AI is fundamentally different from a laboratory demonstration.
A knitting hall is a demanding environment for electronics and software: heat, lint, vibration, continuous machine operation, intermittent connectivity and limited access to technical personnel.
Knit-i is therefore built as an edge-AI system, with computation taking place on the machine rather than depending on a remote cloud connection.
The system is built using Intel technology and OpenVINO and is listed in Intel’s Edge AI Catalog.
The company’s view is that industrial AI ultimately has to be judged not only by model accuracy, but by whether the complete system can remain reliable in production.
“With today’s tools, building an AI pilot has become comparatively easy. What breaks when a deployment goes from ten sites to several hundred is almost never the model. It is installation, monitoring, servicing, spare parts, uptime and support over the life of the equipment.” — Harsha Vardhan, Founder & CEO, CountAI
From Knit-i to a broader textile AI platform
Knit-i is CountAI’s largest deployed product, but the company is extending its computer-vision technology across multiple stages of textile manufacturing.
Its portfolio includes systems for knitted fabric, woven fabric, printed and dyed fabric, cut pieces, yarn cones and final fabric inspection.
The company is also developing a factory intelligence layer that can connect inspection data with production, machine and ERP information, allowing manufacturers to understand quality and production performance across the plant.
The long-term ambition is to move textile quality inspection from isolated inspection points towards a connected, data-driven quality infrastructure across the factory.
Building from India
CountAI was founded in 2022 and has built its business without external investment.
Its systems are deployed in textile manufacturing environments across multiple countries, with the company initially building its installed base in India and expanding internationally.
The company believes the opportunity extends well beyond automating inspection.
As manufacturers face tighter quality requirements, increasing pressure on margins and difficulty finding skilled inspection personnel, the ability to detect and respond to quality problems at the point of production is becoming increasingly important.
“The most valuable thing about being in this ecosystem is not the technology. It is that when a large manufacturer asks who stands behind us, we no longer have to answer that question entirely on our own.” — Harsha Vardhan, Founder & CEO, CountAI
What comes next
For CountAI, the Intel award is less a destination than a validation of a broader thesis: AI for manufacturing has to leave the screen and become part of the machine.
The company is now focused on expanding its deployment footprint, working with textile manufacturers and industry partners, and extending its inspection technology across more stages of textile production.
As edge computing becomes increasingly capable, CountAI believes the next generation of manufacturing AI will not simply report what happened on the production floor.
It will help machines understand what is happening while production is still taking place — and act on it.
About CountAI
CountAI is an Indian industrial AI company building computer-vision and edge-AI systems for textile manufacturing. Its flagship product, Knit-i, provides real-time quality inspection on circular knitting machines, allowing manufacturers to detect defects as fabric is being produced. The company is extending its technology across yarn, knitted and woven fabric, printing, cut-piece and finished-fabric inspection, alongside factory-level intelligence solutions.
Founded in 2022, CountAI is headquartered in India and is an Intel Edge AI partner listed in the Intel Edge AI Catalog.
Website: counton.ai