Voice-Powered AI Startup Voinosis Wins Seoul Pitch Competition to Secure Coveted Spot at TechCrunch Disrupt

Each year, TechCrunch’s flagship Startup Battlefield 200 competition serves as a global launchpad for hundreds of early-stage companies striving for the opportunity to pitch their innovations directly to venture capitalists and the international technology community at the TechCrunch Disrupt conference. Securing a slot in the Battlefield 200 can dramatically alter the trajectory of a young company, offering unprecedented visibility, investor access, and validation on the world stage.

The competitive path to the global stage intensified this week in Seoul, where nine promising early-stage South Korean startups took the stage at DETO 2026, a prominent technology conference held in South Korea’s capital. These companies competed fiercely for a single, guaranteed golden ticket into next year’s Startup Battlefield 200.

The competing companies operate across a wide array of high-stakes deep tech sectors, including healthcare, financial technology, sustainable energy, advanced semiconductors, and artificial intelligence. Bound by ambitious international growth strategies, all nine contenders share a common objective: successfully entering the competitive United States market as a cornerstone of their broader global expansion plans.

After a rigorous evaluation of their technologies, market potential, and pitch presentations, Voinosis emerged as the winner of the Seoul competition. Voinosis leverages artificial intelligence to detect early changes in brain health through subtle voice analysis. While Voinosis claimed the top prize and direct entry, the remaining eight startups showcased equally compelling innovations tailored to address complex industrial, medical, and technological challenges across the globe.

Voinosis

Taking the crown at the Seoul preliminary event, Voinosis is developing a sophisticated artificial intelligence tool designed to analyze fundamental speech patterns—such as pitch fluctuations, pauses, rhythm variations, and response times—to detect physiological and neurological changes linked to brain health. Notably, the entire screening process takes less than five minutes per user, offering a rapid and non-invasive approach to preliminary neurological assessment.

A significant hurdle for voice-based diagnostic tools has traditionally been language dependence. However, Voinosis reports that its technology has already undergone testing in Greek, Japanese, and English, demonstrating that its product functions effectively without requiring separate, exhaustive training for each individual language. In a clinical study involving 146 participants, the startup claims its technology achieved an impressive 88 percent accuracy rate in screening for potential cognitive issues.

Voinosis has built a diverse go-to-market strategy targeting hospitals, senior living facilities, insurance providers, and individual consumers. Looking ahead, the startup plans to seamlessly integrate its voice check assessments with comprehensive cognitive training regimens, formal evaluations, and specialized treatments administered by licensed healthcare professionals.

The company was founded by Grace Shin, who brings more than 25 years of hands-on medical experience as an otologic surgeon. Drawing on her extensive clinical background, Shin is guiding the company through its current fundraising phase. Voinosis is actively raising a Series A investment round and has already secured a prominent South Korean pharmaceutical company as a strategic investor to accelerate its growth and market penetration.

Awesome Lab

Based in Seoul, Awesome Lab is tackling the critical global challenge of energy efficiency and carbon reduction with its flagship innovation, ThermoFlux. ThermoFlux is an advanced heating system that utilizes water electrolysis and ionization to generate thermal energy. The versatile system is engineered to efficiently manage heat production for large buildings, municipal water supplies, and intensive industrial manufacturing processes.

As the world increasingly turns toward sustainable technologies to lower carbon footprints, Awesome Lab highlights its system’s capability to drastically reduce both overall energy consumption and harmful carbon emissions. To execute its international expansion effectively, Awesome Lab is tailoring its go-to-market strategy to fit the nuances of each geographic region. Specifically, in the United States, the company plans to supply core original equipment manufacturer modules while actively partnering with established boiler, water heater, and heating, ventilation, and air conditioning companies to co-develop next-generation products.

Crosshub

Navigating foreign travel, identity verification, and cross-border payments often presents significant friction for international visitors. Crosshub addresses this friction with its proprietary platform, IDBlock, which ingeniously combines rigorous identity verification with secure payment services. The underlying architecture of the platform relies on self-sovereign identity and zero-knowledge proof technologies.

By implementing these advanced cryptographic methods, Crosshub enables travelers to prove their identities across international borders without being forced to share unnecessary personal data, thereby enhancing privacy and security. Furthermore, Crosshub is integrating this identity framework directly with financial payment systems, creating a streamlined ecosystem designed to dramatically improve access to essential financial services for international travelers and mobile populations.

DeepFusion AI

For the fast-evolving sectors of autonomous mobility, robotaxis, maritime surveillance, and defense operations, situational awareness is paramount. DeepFusion AI has engineered a sophisticated 4D radar-based imaging system outfitted with a complementary array of advanced sensors tailored precisely for these demanding environments.

The system ingeniously combines radio frequency sensors with traditional radar and lidar technologies. This multi-sensor fusion allows the platform to accurately detect objects, track movement, and calculate precise distances even in conditions where there is no direct line of sight. Notably, DeepFusion AI packs this comprehensive suite of sensors into a single, highly customizable hardware unit, offering system integrators a versatile and robust solution for complex autonomous navigation and surveillance tasks.

EFLOW

The proliferation of advanced robotics, unmanned aerial vehicles, commercial drones, and low-earth-orbit satellites has created an urgent demand for more efficient actuation and propulsion systems. EFLOW addresses this engineering bottleneck by developing specialized electric motors capable of producing significantly greater torque—the rotational force required for heavy lifting and precise movement—while maintaining a remarkably compact physical footprint.

Beyond core propulsion motors, the startup is also heavily invested in developing advanced actuators. These specialized components act as the vital muscles and tendons of robotic systems, carefully controlling how robotic joints bend, rotate, and respond in real-time operational environments.

Otiton Medical

Medical diagnostics require precision, speed, and reliable data. Otiton Medical addresses these needs through the development of two distinct medical devices designed for primary care and specialized examinations. One of the company’s flagship devices specializes in high-precision body temperature measurement, while the second is engineered to capture high-resolution images of the ear canal and perform detailed analyses of the eardrum.

To ensure diagnostic reliability, Otiton Medical trained its proprietary artificial intelligence on an extensive database comprising approximately 55,000 distinct images of human eardrums. This rigorous training dataset has paid off, as the company’s diagnostic device has successfully secured official medical device certification within South Korea, clearing the path for broader clinical adoption.

Riibotics

Deploying automation technology within modern warehouses is often hindered by lengthy setup times, exorbitant deployment costs, and complex system integrations. Riibotics claims it can radically disrupt this paradigm by deploying fully operational robots into warehouse environments within a single day.

The startup aims to drastically slash capital expenditures and deployment overhead for warehouse operators by streamlining every phase of the setup process. This streamlined methodology encompasses comprehensive infrastructure setup, meticulous warehouse mapping, custom system integration utilizing specialized middleware, precise task programming, and rigorous safety checks to ensure a smooth transition to automated operations.

Ternell

Semiconductor innovation remains the bedrock of modern artificial intelligence advancements. Based in industrial Ulsan, South Korean semiconductor startup Ternell is developing a fundamentally new type of computer memory architecture specifically engineered to help AI chips consume significantly less electrical power.

Traditional memory systems rely on a conventional two-state configuration to store digital data. In contrast, Ternell’s innovative T-SRAM system introduces a third state, enabling much more efficient data storage per unit of area. The company was founded in 2019 by KyungRok Kim, a former engineer at Samsung Electronics. Rigorous internal tests conducted with 28-nanometer chips revealed remarkable performance metrics, showing a 33 percent to 102 percent increase in memory capacity per unit area, alongside a dramatic 55 percent to 71 percent reduction in the energy required to access stored data when compared directly with conventional SRAM architectures.

Testbank

Rounding out the cohort of nine innovative contenders is Testbank, an educational technology startup specializing in the digital transformation of static study materials. Testbank has developed an innovative application named Solve, which is designed to seamlessly convert printed educational assets—such as heavy school workbooks and comprehensive test-preparation books—into fully interactive digital versions.

The startup harnesses advanced artificial intelligence algorithms to accurately identify question and answer spaces within publishers’ original PDF files while faithfully preserving the original visual layout and typography. Through the Solve platform, students can write their answers directly onto the digital page and receive automated grading feedback instantly.