SELVAS Al

What SELVAS Al does

SELVAS AI (KOSDAQ: 108860) is a South Korea-based artificial intelligence company founded in 1999 and listed on KOSDAQ since 2009, headquartered in Gasan Digital Innovation Park, Seoul. The company develops deep-learning-based AI technologies spanning speech recognition (STT), text-to-speech (TTS), handwriting recognition (HWR), optical character recognition (OCR), natural language processing (NLP), and disease-onset prediction. Core technology includes a proprietary End-to-End (E2E) speech recognition engine, deepTTS synthesis with emotional expression and on-device deployment, and an RNN-LSTM-based disease prediction model. Its product portfolio is anchored by the Selvy brand (STT, TTS, OCR, Pen, MediVoice, Note, Checkup) plus higher-level domain products — Selvy MediVoice for medical documentation, SELVAS Note for investigation/government recording, SELVAS SpeechEdu for language education, Selpub for digital textbooks, and Odiro for AI audiobook production.

The company operates a vertically-segmented go-to-market motion across healthcare (Samsung Seoul Hospital, Busan Baek Hospital, Kangbuk Samsung Hospital, Korea National Cancer Center, Ministry of National Defense), finance (Woori Bank, KB Kookmin Bank, Woori Card, Hana Life Insurance, Samsung Life Insurance), public sector (Ministry of Justice KICS system, Daejeon Fire Department, Korean Coast Guard, National Assembly), education (Selpub digital textbook pilots), mobility (vehicle voice middleware for Amazon Alexa/Cerence/Baidu/SoundHound), and content (YES24, Millie's Library, Kyobo, Ridi, Aladin, Munpia e-book and web-novel platforms). Its subsidiary structure includes Selvas Healthcare (AI medical services), Mediana (acquired January 2024 — patient monitoring, AED, automatic CPR machines), and the SelvasInvision joint venture (February 2024, with US-based UltraSight for AI cardiac ultrasound).

Revenue is generated primarily through enterprise licensing and subscription contracts (Selvy MediVoice, Selvy Checkup API/built/rental, SpeechEdu API), one-time government procurement licenses, API usage fees for TTS in e-book platforms, hardware sales through Mediana, and royalty income from automotive voice-middleware licensing. Customers number 520+ across digital transformation engagements, with 200+ commercialized cases and reported 2024 consolidated revenue of 1,124.5 billion KRW (Korean notation; approximately $86M USD) representing an all-time high. The company has raised approximately 135.9 billion KRW via two 2023 rights offerings to fund R&D and new business expansion.

Name

SELVAS Al

Legal name

주식회사 셀바스에이아이

Website

https://selvasai.com

Company type

Public

Founded year

1999

Operating status

Operating

Headcount range

11–50 employees

Short description

SELVAS AI is a Korea-based AI company developing speech recognition, text-to-speech, handwriting recognition, OCR, NLP, and disease-prediction technologies, serving 520+ enterprise and public-sector customers across healthcare, finance, government, education, and mobility.

Ownership category

akta.pro rank

Product category

AI Speech Recognition Software

NAICS

Custom Computer Programming Services (541511), Computer Systems Design and Related Services (5415)

SIC

Services-Computer Programming Services (7371)

akta.pro primary industry

On-Device Speech & Audio AI (wake word, ASR, enhancement) (HDAAAJAF)

akta.pro secondary industries

Pronunciation, Speech Recognition & Accent Training Tools (EDAFAOAG), Language Assessment, Placement Testing & Proficiency Certification Tools (EDAFAOAH), Fine-Tuning, Adaptation & Custom Model Training (PEFT/LoRA/RLHF) (HDAAACAC)

Keywords

• AI speech recognition
• Text-to-speech synthesis
• Medical voice documentation
• Handwriting recognition AI
• Disease prediction models

Headquarters

HQ city

Seoul

HQ country

South Korea

HQ region

Asia

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Operations, Marketing or Sales, Infrastructure

Revenue model

1. AI Solutions (STT, TTS, OCR, NLP): Licensing and API-based revenue from AI speech recognition, synthesis, handwriting recognition, and NLP technologies to enterprise customers
2. Healthcare AI Products (Selvy MediVoice, Selvy Note): Sales of AI medical voice recording solutions to hospitals and healthcare providers, deployed via server, cloud, or API models
3. AI Education Products (Selvas SpeechEdu, Selpub): AI language learning evaluation and digital textbook platform services for educational institutions
4. Disease Prediction (Selvy Checkup): Disease prediction analysis services provided to insurance companies, health management services, and medical institutions via API, building, or rental models
5. Public Sector Solutions: Voice recognition and recording solutions for government agencies including police, fire departments, courts, and parliamentary bodies
6. Mobility AI Solutions: Voice recognition middleware for automotive and mobility applications, generating royalty income from technology licensing
7. Mediana Medical Devices: Connected medical device business including patient monitoring systems, AED, and automatic CPR machines through subsidiary Mediana

Pricing tiers

Go-to-market motion3 records

Distribution channels6 records

Marketing channels5 records

Core offering

SELVAS AI develops and commercializes deep-learning-based AI technologies — speech-to-text (STT), text-to-speech (TTS), handwriting recognition (HWR), OCR, NLP, and disease-onset prediction — packaged into standalone products and enterprise solutions. Its offerings are deployed via server installation, cloud, or API delivery into healthcare (Selvy MediVoice, Selvy Checkup), education (SELVAS SpeechEdu, Selpub, Selvy Pen), finance (AI Navigator for bank kiosks), government (Selvy Note, 119 emergency response, maritime response system), and mobility (vehicle voice middleware).

Product overview

SELVAS AI provides a comprehensive AI technology platform built around speech and language intelligence, with a portfolio of both core AI solutions and consumer-facing products. The core AI Solutions (Selvy STT, Selvy TTS, Selvy Pen, Selvy OCR, NLP, Selvy Checkup) serve as foundational technology enablers. On top of these, SELVAS AI has built domain-specific products across healthcare (Selvy MediVoice for medical documentation, Selvy Checkup for disease prediction, Central Monitoring Solution for hospital-wide patient monitoring), education (SELVAS SpeechEdu for speaking/writing evaluation, Selpub for digital textbook creation, Selvy Pen for handwriting), finance (voice interfaces for banking kiosks and contact centers), mobility (vehicle voice middleware, railway AI), public safety (119 emergency response, police investigation systems), and entertainment (Odiro for audiobook production, Selvy Singing Voice). The company also operates subsidiary brands including Selvas Healthcare (Braille devices), Mediana (medical devices), SelvasHealthCare, Selpub, and OnFit.

Differentiator

Problem solved

Functional benefit

Brands

• Selvy (셀비): Primary product brand for speech recognition (STT) products including Selvy STT, Selvy Note, Selvy MediVoice, Selvy Checkup, Selvy OCR, Selvy Pen
• Selvy deepTTS
• SELVAS MediVoice
• SELVAS Note
• SELVAS SpeechEdu
• SELVAS Alrimi
• Selpub
• odiro
• Selvy Checkup
• OnFit
• DIODICT

Products and services

• Selvy STT (Speech-to-Text) AI Speech-to-Text technology with an in-house E2E (End-to-End) speech recognition engine supporting Korean, English, Chinese, and Japanese. Available cloud-based or on-device. Sold as an API service to enterprise customers and as the foundation for Selvy MediVoice and Selvy Note.
• Selvy TTS (Text-to-Speech / deepTTS) Text-to-speech voice synthesis including deepTTS (premium cloud TTS with emotional expression and neural vocoder) and deepTTS On-Device (offline TTS SDK for Android/iOS/Raspberry Pi OS). Supports 8 languages with 22 voices; sold via cloud, on-device SDK, and as embedded OEM voice for e-book and kiosk platforms.
• Selvy OCR Optical character recognition technology with 93.6% accuracy on mathematical formula recognition. Applied to ID cards, credit cards, and mathematical documents. Distributed via the Hyphen Data Marketplace and as an on-device solution for e-book and document conversion.
• Selvy Pen (Handwriting Recognition) Handwriting recognition technology for mathematical equation recognition (math solution processes and answer verification) and Korean script stroke-order recognition (dictation). Applied in educational settings for math and Korean-language learning.
• Selvy Checkup AI disease-onset probability prediction service that analyzes health-examination data via RNN-LSTM deep learning models and predicts 10 major diseases (diabetes, heart disease, stroke, dementia, and 6 cancers) over a 4-year horizon with 93% self-evaluated accuracy. Available via API service, built service, or rental service; applied at insurance companies, health checkup centers, and healthcare services.
• Selvy MediVoice AI medical voice-recognition solution using a medical-specific sLLM to recognize and analyze doctor-patient conversations in real time, automatically summarizing them in SOAP format. Integrates with hospital EMR systems. GS Certification Grade 1. Deployed in over 50% of major Seoul superior general hospitals including Samsung Seoul Hospital, Samsung Medical Center, Kangbuk Samsung Hospital, and Korea National Cancer Center.
• SELVAS SpeechEdu Integrated AI language-education platform combining speaking evaluation (SpeechBox: STT-based dictation, sentence pronunciation rating, phonics) and writing evaluation (SentenceBox: essay missions, descriptions, answer evaluation) on a single platform. Supports Korean, English, Chinese, and Japanese. Delivered API-based to educational institutions to reduce initial build costs.
• Selvy Note AI voice-documentation solution for investigations, consultations, and meetings. Converts speech to text with speaker separation, real-time transcription, and AI summarization. GS Certification Grade 1. Selvy Note 3.0 applies the next-generation E2E speech recognition engine and supports file- and folder-level batch voice recording for investigative/government institutions.
• Selvy Pen Guide AI handwriting-correction solution for digital Korean language education. Recognizes Korean character stroke order and evaluates handwriting quality, helping students improve penmanship through AI-powered feedback.
• Odiro AI voice-content production service for creating audiobooks and voice content, including an AI Audiobook Editor that reduces audiobook production time by up to 90%. Features 25 style-specific AI voices for genre fiction including romance and fantasy. Uses deepTTS technology and supports offline audiobook listening.
• Selvy deepTTS On-Device On-device AI text-to-speech SDK deployable inside smartphone apps for offline use. First on-device AI TTS commercialization in Korea. Supports Android, iOS, and Raspberry Pi OS with no network connection or text-length limit. Reduces server operations and data-transmission costs.
• Selpub Customized e-book authoring, learning, and management platform for teachers. Converts PDF, PPT, Word, HWP files into digital textbooks. Includes AI handwriting recognition (math/Korean), AI voice intelligence (pronunciation evaluation, reading-aloud), and AI content recommendation. Composed of Selpub Creator, Selpub Viewer, and Selpub Cloud. AWS Partner Path certified.
• OnFit Smart fitness-center management platform covering member management, workout tracking, and integrated fitness-center operations.
• DIODICT Multi-language electronic dictionary with 30+ languages and 200+ content types. Mobile-platform dictionary application (iOS service ended January 1, 2023).
• Selvy LRS (Learning Record Store) Learning-activity data management solution tracking learner engagement with educational content. Officially registered on Korea Public Procurement Service (조달청) Digital Service Mall. Provides student and teacher dashboards showing learning-tool activity, content usage, and quiz results. Holds ADL (xAPI) certification.
• Selvy SR API AI English pronunciation-learning API based on speech-recognition technology, available as a subscription-based API service. Enables English sentence pronunciation evaluation and learning through an accessible API for educational and EdTech customers.
• Selvy Singing Voice AI singing-synthesis technology capable of producing singing AI voices that mimic singing styles, lyrics, melody, and timbre. Applied to entertainment-content creation.
• Vehicle Voice Recognition Middleware Voice-recognition middleware for vehicles providing multilingual speech recognition (Korean, English, Chinese, Japanese, European languages) with mass-production experience at major Korean automakers and expansion planned into railways and industrial mobility.
• BrailleSense 7 Braille terminal with on-device Google Gemini AI integration developed by affiliate Selvas Healthcare. Features in-house finger-sensing Active Responsive Braille Display, providing simultaneous visual and tactile output across three models tailored for educational and office environments.

Quantifiable outcome

• Medical record documentation time reduced to 1/3 of original (reported at Samsung hospitals)
• +4 more outcomes

Named customers16 records

Segments6 records

Ideal customer profiles6 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration17 records

AI capability14 records

Feature8 records

Partnerships

19 partnerships are on record, tiered core and minor.

• LG CNScoreTechnology or Integration · 26 March 2026Integration of LG CNS AI technology for analyzing customer query intent and searching bank's data platform, combined with SELVAS AI's STT/TTS for Woori Bank smart teller machines.
• Amazon AlexacoreTechnology or Integration · 24 March 2026Collaboration for vehicle voice recognition middleware development. SELVAS AI is upgrading its voice recognition middleware technology for vehicles through partnerships with global voice recognition companies including Amazon Alexa, Baidu, Cerence, and SoundHound. Multilingual support for Europe, America, and Asia.
• BaiducoreTechnology or Integration · 24 March 2026Collaboration for vehicle voice recognition middleware with multilingual support for global automotive markets.
• CerencecoreTechnology or Integration · 24 March 2026Collaboration for vehicle voice recognition middleware with multilingual support for Europe, America, and Asia automotive markets.
• SoundHoundcoreTechnology or Integration · 24 March 2026Collaboration for vehicle voice recognition middleware with multilingual support for global automotive markets.
• Mediana (메디아나)coreStrategic or Co-development Partner · 23 March 2026Subsidiary partnership for AI-based smart ward solution development. Mediana provides patient monitoring data and device integration. Combined with SELVAS AI's AI analysis and voice recognition for integrated patient monitoring.
• Easy Care Tech (이지케어텍)coreStrategic or Co-development Partner · 23 March 2026Joint development of AI-based smart ward solution. EasyCareTech builds EMR integration system and user interface. Consortium plans to expand into overseas markets including Middle East leveraging EasyCareTech's existing experience supplying hospital information systems to Saudi Arabia.
• SELVAS Healthcare (셀바스헬스케어)coreStrategic or Co-development Partner · 17 March 2025Subsidiary partnership at KIMES 2025 showcasing AI convergence connectivity solutions combining medical devices with AI/software technologies.
• Ministry of National Defense / Armed Forces Medical Command (국방부/국군의무사령부)coreStrategic or Co-development Partner · 30 October 2023Supply of AI medical solutions for 'Advanced ICT-based Smart Medical System' project. Planned application at military capital hospitals in the first half of next year.
• Daegyo (대교)minorStrategic or Co-development Partner · 20 July 2023MOU with Daegyo Definition for AI digital textbook education innovation. Collaboration on textbook content digitalization technology.
• UltraSight (울트라사이트)coreStrategic or Co-development Partner · 8 May 2023Partnership for AI cardiac ultrasound solutions. UltraSight provides AI guidance technology enabling non-specialist medical professionals to perform accurate heart ultrasound examinations. Collaboration to expand into remote medical services and Korean/Asian markets. A joint venture 'SELVAS Invision' was established in 2024 for AI cardiac ultrasound business.
• LG U+ (LG유플러스)coreTechnology or Integration · 5 April 2023Expansion of voice intelligence-based conversational AI services targeting major telecom companies. Combined SELVAS AI voice recognition and synthesis with LG U+ AI chatbot solutions.
• Robotics (로보티즈)minorStrategic or Co-development Partner · 4 April 2023MOU for conversational AI robot business. Collaboration to develop GPT-applied conversational AI robot solutions.
• WJS (위지윅스튜디오)minorStrategic or Co-development Partner · 29 March 2023Strategic investment and partnership for AI digital twin content business.
• Meracler (메라커)minorStrategic or Co-development Partner · 29 March 2023Joint development and business cooperation for AI-based digital twin content.
• Samsung Medical Center (삼성서울병원)coreStrategic or Co-development PartnerAI smart hospital co-development. SELVAS AI applied voice recognition technology across all medical departments at Samsung Seoul Hospital, developing solutions for radiology, emergency, surgery, and nursing documentation.
• Chungnam National University Hospital (충남대학교병원)minorStrategic or Co-development PartnerMOU for joint research in healthcare technology. Collaboration with SELVAS Healthcare for new technology research.
• Woori Bank (우리은행)coreTechnology or IntegrationVoice technology application to Woori Bank's smart kiosk enabling customers to perform financial transactions via voice commands through AI Navigator system.
• Hyunso (현소)minorOthersAI voice recognition technology application in elevator emergency response systems with Korea Elevator Safety Agency.

Scale indicators9 records

Recent moves7 records

Expansion highlights7 records

Broad incumbents

• Nuance Communications (Microsoft): Microsoft-owned speech AI leader specializing in healthcare documentation (DAX), automotive, and enterprise voice. Most direct global competitor to Selvy MediVoice in medical speech recognition and represents the largest commercial threat to Selvas AI's hospital vertical.
• Kakao Enterprise: Korean conglomerate AI arm offering KoGPT, speech, and chatbot services. Competes with Selvas AI in Korean enterprise and financial-sector conversational AI deployments.
• iFlytek: China's leading speech recognition and AI company with deep specialization in Mandarin, education, and healthcare verticals. Closest analog to Selvas AI in being a vertically-integrated, non-English speech AI specialist with education and government exposure.
• Naver Cloud (HyperCLOVA): Korean hyperscaler with HyperCLOVA LLM and CLOVA Speech APIs. Most direct Korean competitor across enterprise STT/TTS, with deeper capital base and broader platform play threatening Selvas AI's domestic enterprise wins.

Direct peers

• SoundHound AI: Voice AI company with Houndify platform serving automotive, restaurant, and IoT. Direct competitor in vehicle voice middleware and conversational AI, where Selvas AI is also pursuing multilingual automotive deployments.
• Cerence Inc. Pure-play automotive voice AI company spun out from Nuance. Direct competitor in vehicle voice recognition middleware where Selvas AI is partnering with Amazon, Baidu, Cerence, and SoundHound — competing for the same automaker design wins.

Emerging players

• Voiceflow: Conversational AI platform for designing and deploying voice/chat assistants. Adjacent comparable for Selvas AI's AI Navigator financial kiosk and contact center conversational AI offerings, though Voiceflow focuses on design tooling rather than core ASR/TTS engines.
• Sensory Inc. On-device voice and edge AI specialist focused on wake-word, speech recognition, and biometrics on embedded devices. Direct comparable to Selvas AI's deepTTS On-Device and Selvy STT On-Device offerings for mobile and IoT.
• LumenVox: Speech recognition and voice biometric technology provider for contact centers and enterprise. Comparable in offering STT/TTS APIs to mid-market and enterprise customers, though with less specialization in Korean and medical domains.

Regional players

• DeepBrain AI: Korean AI company specializing in AI humans, synthetic video, and TTS for enterprise and education. Overlaps with Selvas AI's deepTTS and Odiro audiobook voice generation capabilities in Korean-language voice synthesis.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat6 records

Key risks6 records

Key highlights6 records

Customer concentration

• Facebookfacebook.com/diotek
• LinkedInlinkedin.com/company/selvas-ai-inc.

Compliance6 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles4 records

Subsidiaries4 records

Funding overview

Funding rounds2 records

Investors2 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments1 record

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about SELVAS Al