AI for Swim Verification: How ZLiX Tracks Progress for Beginning Swimmers
Bulgarian swimmer and technology developer Valentin Milkov is building a digital swimming passport that verifies skills, awards certificates and helps families see real progress.
Val was a fast swimmer and was coached by at Indiana University by Kris Kirchner and later, while redshirted due to a shoulder injury, he assisted Ray Looze as a coach.
Swimming instruction often depends on information that stays inside the coach’s head. A coach knows when a child has improved, when a skill has become reliable and when the swimmer is ready to advance. Parents and guardians may see attendance, effort and enthusiasm, but they do not always receive a clear record of what the swimmer can actually do.
Valentin Milkov wants to make that progress visible.
Milkov is a Bulgarian swimmer, technology project manager and developer of ZLiX, an application designed to verify swimming skills through uploaded video. The system follows beginning swimmers through a progression of tasks, confirms whether they have completed each skill and awards digital certificates or badges as they advance.
During a recent interview for Heavy Or Not, The O.G. Swim Guide, Milkov described Zlix as a type of digital swimming passport.
The goal is straightforward. A swimmer, parent or coach records a required skill, uploads the video and receives an assessment. The swimmer either passes the task or receives guidance about what needs improvement.
Zlix does not attempt to replace the coach. It gives coaches and families a common record of the swimmer’s development.
Swimming Led Milkov Around the World
Milkov is based in Bulgaria, but swimming has taken him far beyond his home country.
He attended Indiana University, drawn by the program’s history and its connection to Olympic champion Mark Spitz and legendary coach Doc Counsilman. Milkov says he became a U.S. Open champion while swimming at Indiana.
He later completed a master’s degree in England and remained involved in competitive swimming. He also built a professional career managing technology, maritime and information-technology projects.
Swimming never disappeared from his life. It stayed in the background as a field where he wanted to contribute.
Around 2015, Milkov began discussing artificial intelligence and swimming with friends working at the leading edge of AI for media. Their early concept was called SwimMind. The idea focused on using artificial intelligence to provide feedback based on recognized swimming practices, especially for beginning swimmers.
The concept earned recognition through a technology initiative connected with major Olympic and corporate organizations. The award confirmed that the idea had promise, but the technology still faced a difficult environment.
Water creates problems for computer vision. Reflections, bubbles, camera angles, lighting and the movement of the surface make reliable video analysis harder than analysis performed on land.
That challenge remains.
Zlix Begins With Verification
Milkov waited for artificial-intelligence tools to improve and eventually developed the front-end Zlix application himself.
The current system is intentionally simple. It is a minimum viable product built to test interest and gather useful information from actual users.
Zlix currently focuses on verification, not advanced stroke analysis.
That distinction is important.
The application does not claim to analyze every detail of a swimmer’s technique or act as an automated replacement for a skilled instructor. It checks whether the swimmer completes a clearly defined task.
A task might require a swimmer to:
- travel 10 meters with flotation support;
- travel 10 meters without arm floats;
- hold a front or back starfish position for five seconds;
- demonstrate a streamline;
- complete a specified distance in freestyle or backstroke; or
- finish a more advanced swimming assignment.
The progression includes 13 levels. It begins with introductory levels represented by aquatic characters such as the seahorse and continues through increasingly demanding skills. The highest current level reaches a 400-meter individual medley.
The emphasis remains on the path from beginning swimmer to capable swimmer. Zlix is not presently designed as a technical-analysis system for elite athletes.
The Swimmer Uploads a Video
The user experience is deliberately uncomplicated.
The swimmer selects a level and reads the required task. The swimmer, parent or coach records the performance and uploads the video. Zlix then evaluates whether the task has been completed.
Some skills already produce strong automated results. Milkov reports approximately 90 percent accuracy for certain clearly defined actions, such as making a star shape in the water.
Ninety percent is not enough for a system that awards certificates and communicates directly with families. Zlix therefore keeps people involved in the review process.
Human reviewers currently work behind the system to confirm assessments and prevent swimmers from receiving an incorrect result. Milkov wants the swimmer’s experience to be dependable, especially during the early stages of the product.
This combination of artificial intelligence and human oversight allows Zlix to improve without pretending that the technology has already reached perfect accuracy.
A Digital Certificate Makes Progress Visible
The certificate is a central part of the system.
A swimmer who passes a level receives a digital record of the accomplishment. Some levels contain several tasks, so the swimmer may need to complete more than one demonstration before earning the full certificate.
The certificate gives the swimmer a sense of achievement. It also gives the family something concrete.
A parent can see that three weeks ago the swimmer completed only a few introductory skills. Later, the same child may have added a streamline, a float, a kick, a breathing pattern and a full length of a stroke.
That progression is more informative than attendance alone.
A child does not improve simply because the child has attended six lessons. Improvement appears through new abilities that can be demonstrated, verified and recorded.
The digital badges also give coaches another way to motivate younger swimmers. Aquatic characters, levels and certificates can turn skill development into a visible journey instead of an unclear sequence of lessons.
Different Programs Can Use Different Progressions
Swimming schools do not all follow the same curriculum.
Some use numbered levels. Others use letters, animals, superheroes or locally designed achievement systems. Swim America, Swim Australia and independent schools organize their programs differently.
Zlix can accommodate that variation.
Milkov wants to maintain a standard public version of the program based on established practices from countries including Australia and the United States. At the same time, he says a swimming school can work with Zlix to create a customized version using its own:
- curriculum;
- level names;
- logos;
- skill requirements;
- visual identity; and
- certificates.
A school could therefore preserve its existing teaching system while using Zlix for video submission, verification and digital recognition.
That flexibility may become one of the application’s strongest features. The technology does not require every swimming organization to abandon its own teaching language.
Zlix Is Intended to Remain Free for Swimmers
Milkov says the end-user application is intended to remain free.
Parents, swimmers and individual users would not pay to upload their performances and receive verification. The initial business model relies on the certificates and potential sponsorship connected with them.
As the platform grows, certificates could include sponsor participation or related commercial support. Customized systems for swimming schools may also create additional opportunities.
Artificial intelligence is not free to operate. Video storage, data processing, machine learning and continued model development require money. Milkov is testing whether enough swimmers, families and coaches value the service to justify further investment.
More use also gives the system more examples from which it can learn. Each properly handled video can help improve future performance, provided privacy, consent and data practices remain clear.
Milkov also hopes that growing public interest may help Zlix qualify for European Union development funding.
Zlix Is Not Another Elite Stroke-Analysis Platform
Much of the artificial-intelligence work in swimming begins at the competitive end of the sport.
Other systems use cameras to count strokes, measure speed, calculate turn times, assess starts and produce race data across several lanes. Those tools serve competitive teams and high-performance swimmers.
Zlix deliberately starts elsewhere.
It focuses on the swimmer who is learning to float, streamline, breathe and travel safely through the water.
Milkov believes that beginning swimmers receive less technological attention even though their development is critical. Families often struggle to understand the curriculum, and standards vary widely between schools and countries.
Zlix attempts to create a visible progression without turning every lesson into an advanced biomechanics session.
The difference can be summarized clearly:
Stroke-analysis systems examine how a swimmer performs a movement. Zlix currently verifies whether the swimmer completes the assigned skill.
Coaches Stay in the System
Milkov does not present Zlix as a substitute for swimming teachers.
The coach still teaches the skill, organizes practice, corrects the swimmer and decides how learning should progress. Zlix provides documentation and an additional verification layer.
This approach supports a partnership among the swimmer, coach and family.
The instructor can assign a skill. The swimmer can practice it. A parent can record the attempt. The application can verify the result. Everyone can see the same achievement record.
That shared record may be especially useful when swimmers change instructors, programs, cities or countries. A digital swimming passport can travel with the swimmer.
Future Development Could Become More Interactive
Milkov sees a much larger future for the platform.
The aquatic characters associated with each level could eventually become interactive guides. A young swimmer might communicate with an AI character, receive encouragement and move from one character or level to the next.
More advanced versions could provide closer technical feedback. A future facility might combine underwater cameras, above-water cameras and AI-supported coaching stations.
Milkov imagines a swimming pool where each lane has access to an AI assistant. A swimmer could finish a repeat, touch the wall and receive immediate information about breathing, stroke execution or another assigned performance measure.
A coach could then supervise more swimmers without losing access to individual data.
That vision requires better cameras, stronger AI models, more training data and careful integration with human coaching. Zlix is not there yet. Milkov is taking the narrower and more practical first step of skill verification.
Collaboration Could Expand the System
The interview also identified several possible directions for collaboration.
A South African coach has already created a collection of instructional drill videos covering the competitive strokes. Those materials could help explain the difference between drills and provide examples for older children.
Developers working on high-performance swimming analysis may also possess trained models that recognize strokes, starts, turns and race actions. Their technology could complement Zlix’s learn-to-swim focus.
Educational projects could add another layer. Young swimmers might use artificial intelligence to record reflections about fear, confidence, improvement and support from teammates or teachers.
Together, these tools could document three different dimensions of learning:
- what the swimmer can perform;
- how the skill is technically executed; and
- how the swimmer experiences the learning process.
Zlix currently concentrates on the first dimension.
That focus gives the application a defined purpose instead of allowing it to become another broad AI product that promises everything and delivers little.
A Practical Starting Point for AI in Swimming
Artificial intelligence in sports often arrives with inflated claims. Zlix is more useful when viewed as a developing tool with a specific job.
It verifies swimming tasks. It records progress. It awards digital certificates. It gives parents and coaches a clearer way to see the swimmer’s journey.
The current product remains early. Some assessments require human review. Not every skill has been fully automated. The long-term business model still needs to prove itself.
Those limitations do not erase the value of the idea.
Swimming instruction needs better ways to show progress, especially for beginners. Families should be able to see what has changed. Swimmers should be able to carry a record of their achievements. Coaches should have tools that strengthen instruction rather than push them aside.
Milkov’s direction is disciplined: begin with verification, work alongside coaches, improve the technology through actual use and expand only when the system is ready.
Technology should not frighten the swimming community. It should help coaches and swimmers work more intelligently.
- Hard work remains part of the sport. Smart work can make that effort count.