[SeoulTech x Donga.com Joint Planning] Seoul National University of Science and Technology (hereinafter referred to as SeoulTech) operates various support programs to aid the growth of startups, including preliminary and early-stage startup packages, maker spaces, and global collaborations. Furthermore, in partnership with Donga.com, it provides global news to assist startups in overseas promotion and expansion, introducing promising deep-tech startups both domestically and internationally.
Sports is a history of records. In sports, records such as scores and passes prove a player's present and prepare them for the future. Detailed indicators evaluate a player's skills and play a crucial role in the strategic planning and scouting decisions of clubs. Recently, not only players but also the general public who enjoy sports want to objectively verify their game data.
Kwon Jung-hyuk, CEO of SPOIT / Source=IT Donga
However, creating such records has required a lot of time and manpower until now. This is because people had to manually review the entire game to select highlight scenes and organize the data manually. To solve this inefficiency, the startup SPOIT emerged. SPOIT offers 'SCOUT BOX,' which uses artificial intelligence (AI) to extract about 30 highlight scenes from a 90-minute game and produce customized videos.
SPOIT is developing a platform that goes beyond sports video production to record and analyze individual player data using vision AI technology and provide AI-based feedback. It aims to help teams find the players they need and help players find teams that match their strengths. IT Donga interviewed Kwon Jung-hyuk, CEO of SPOIT, about the founding motivation, services, and future plans of SPOIT.
Continuing a 30-year soccer career into a startupKwon Jung-hyuk, CEO of SPOIT, who played as a professional soccer player for 16 years, founded SPOIT after concluding his 30-year soccer career. The startup idea originated from personal experiences during his active career. He said, "When preparing to transfer to a European team, sports video services were already systematically established in Europe, where professional sports are developed. In contrast, there was no place in Korea that specialized in producing highlight videos for players, making it difficult due to low video quality."
This led Kwon Jung-hyuk to start providing sports video services in 2019. Initially, AI technology was not incorporated. However, the process of checking and editing 10 to 15 games individually to create a player's portfolio video was inefficient, and he realized the necessity of adopting AI technology. Since last year, SPOIT has begun to actively incorporate vision AI technology into its sports video service, SCOUT BOX.
AI-generated sports video 'SCOUT BOX'
SPOIT SCOUT BOX / Source=SPOIT
The core service of SPOIT, SCOUT BOX, is an all-in-one highlight production system designed by experts and AI together. When a 90-minute soccer game video is input, AI tracking technology automatically analyzes and extracts about 30 key scenes. It recognizes and automatically composes highlights of crucial moments that attract the audience's attention, such as goal scenes and the goalkeeper's super saves.
Kwon Jung-hyuk stated, "Thanks to our proprietary AI technology that individually recognizes and tracks the players, or objects, on the field, the editing process is much faster than traditional methods." Additionally, a scouting video director with experience as a K-League power analyst selects and creates customized videos that emphasize the capabilities of players by position. SPOIT has registered three patents related to sports video processing and has applied for seven more.
SPOIT's AI tracking technology / Source=SPOIT
Currently, the focus is on highlights of the entire game, but individual player tracking features are planned to be introduced next year. This will allow automatic extraction of data such as specific player movements, pass success rates, activity levels, and distances covered. Kwon Jung-hyuk emphasized, "By providing precise tracking, highlight videos, and multi-layered game indicators, it leads to actual performance improvement for players. It enables objective evaluation of players and expands matching opportunities for both teams and players."
The technical complexity of soccer, where 22 players are constantly moving, is high. Therefore, SPOIT is targeting futsal, which is played 6-on-6 and has a relatively simpler environment, to secure data. To this end, SPOIT has acquired two directly operated football centers. Kwon Jung-hyuk stated, "Once the technology development is completed based on the data secured from football, AI tracking technology can be applied to various sports such as basketball and marathons, not just soccer."
Expanding services from professional players to the general publicSPOIT targets both B2B and B2C markets. It provides B2B services for player agencies or clubs and B2C services for players without a team, club members, and youth players who individually request videos. Kwon Jung-hyuk added, "In the past, portfolio videos were only necessary for foreign transfers, but now they have become essential even for domestic transfers. Recently, there has been an increase in cases where student athletes request to keep their videos or submit them as profiles."
Personal highlight video made by SPOIT / Source=SPOIT
In August, SPOIT also launched the SCOUT BOX service as a mobile app. SCOUT BOX app users can easily shoot and share their sports activity videos. By directly selecting important scenes such as scoring situations during a game, a 30-second clip before and after is automatically saved. It also provides functions to merge and share without the need for separate editing. Kwon Jung-hyuk announced, "By lightening individual AI tracking technology, we plan to offer highlight scene extraction features in the app next year."
Furthermore, SPOIT plans to provide AI-based feedback using large language models (LLM) in the future. Ultimately, SPOIT aims for AI to act as a personal coach for users and as an auxiliary tool for scouts. For example, based on accumulated player data, AI might advise, "Your left-foot pass success rate is low, so practice more with your left foot." Additionally, it can help scouts find players that meet specific conditions, such as a 'defensive midfielder with a high pass success rate,' and recommend teams that match a player's strengths. Kwon Jung-hyuk stated, "It will become an intelligent partner that aids human decision-making based on objective data."
Step-by-step expansion strategy… Aiming to dominate the domestic market
Kwon Jung-hyuk, CEO of SPOIT / Source=IT Donga
Entering its fifth year, SPOIT plans to continuously enhance the SCOUT BOX service. The enhancement of the app launched in August is also being pursued concurrently. Kwon Jung-hyuk stated, "We plan to improve the performance of major AI algorithms such as object detection, tracking, situation recognition, and automatic highlight extraction, and expand the scope of service commercialization," adding, "Last year, through the Ministry of SMEs and Startups' TIPS program, we strengthened our development workforce and are focusing on applying vision AI technology to sports videos in various ways."
To this end, SPOIT is actively collecting video data. Notably, it plans to expand the application scope of the smart futsal system, which combines cameras and AI technology. Kwon Jung-hyuk stated, "The smart futsal system provides users with personal videos and record data and creates a new revenue model for futsal field operators. Just as the transition from bare ground to artificial turf was natural, the desire to exercise while leaving videos and data will become a natural trend."
SPOIT has been selected as a collaborative company for Oracle Cloud Infrastructure (OCI). Kwon Jung-hyuk stated, "As the organization grows, we are considering ways to grow the company through selection and focus. Practical support, including Oracle Cloud and research and development funds, has been helpful. In particular, we have had the opportunity to secure technological capabilities based on high-performance computing needed for AI model training."
Lastly, Kwon Jung-hyuk expressed his ambition, "We will create a system where anyone can prove their play with data. Our goal is to become a sports company that connects players and teams, contributes to society, and provides value," adding, "We want SPOIT to become a representative company that establishes a data-driven sports ecosystem in Korea."
IT Donga Reporter Kim Ye-ji (yj@itdonga.com)
ⓒ dongA.com. All rights reserved. Reproduction, redistribution, or use for AI training prohibited.
Popular News