Data Labelers Wanted: Help Build the Future of Boxing 🥊

[devinschumacher]

We are looking for people to help us create a dataset of annotated boxing footage that identifies and classifies every punch thrown - including punch type, target, hand used, and outcome - to train AI models that can automatically analyze boxing matches and give fighters the same advanced analytics currently only available to elite promoter

Every punch you label helps level the playing field. Amateur trainers will have the same analytical tools as elite gyms. Independent researchers can study the sport without corporate gatekeepers. Fighters get better data about their own performance.

Perfect For

• Boxing fans who know the sport and want to give back
• Students looking for relevant experience in data/ML
• Anyone passionate about making boxing more transparent
• No technical background required - if you can use YouTube, you can do thisvs

What You'll Do

• Watch boxing footage and identify punch types (jab, cross, hook, uppercut)
• Label key details of the punch like target (head/body), hand used, etc
• Use simple tools - point, click, select from drop-downs (full training provided)

Whats in it for you

• Direct impact on the boxing community
• Credit in all research papers and applications using the dataset
• Experience with annotation tools and data creation
• Official internship credit from SERP AI
• LinkedIn recommendation from SERP founder Devin Schumacher
• Company LinkedIn inclusion - listed as team member on our company page
• Professional reference letter for any job applications
• Portfolio addition - legitimate project experience for your resume

Your Commitment

• Flexible scheduling - work whenever you want
• Small tasks - annotating a punch takes only a few seconds
• Quality over quantity - we value accuracy over speed
• Time commitment - We want as many people to be able to participate who are interested as possible, but it does take time and resources to train and manage and quality check people's work, so we would prefer to work with people who can consistently contribute more than sporadically or only limited amount.

Why This Matters

For decades, boxing analytics have been locked behind corporate walls, creating a two-tier system that benefits the wealthy and powerful while leaving fighters, trainers, and fans in the dark.

CompuBox is the monopoly for punch statistics, used by major networks like HBO, NBC, and ESPN. But here's the shocking reality: CompuBox offers no API, no public database, and no way for anyone outside their exclusive corporate partnerships to access the data they've been collecting for 40 years.

Independent developers acknowledge that "boxing data is very hard to come by, there is no single source of truth" and are forced to build complex scraping systems just to access basic information.

This creates a massive inequality: elite promoters like Top Rank, Golden Boy, and Matchroom Boxing have access to sophisticated analytics and million-dollar insights, while amateur boxers, independent trainers, and developing fighters are left to guess.

We're fixing this.

SERP AI is teaming up with Boxing Undefeated to help change the world of boxing.

This project aims to create the first truly open, comprehensive boxing dataset - starting with punch recognition and classification that we will use to break down through the boxing world's gatekeepers and democratize the analytics that should belong to everyone who makes this sport possible: the fighters, trainers, and passionate fans who keep boxing alive.

Ready to help change the future boxing?

---

Examples of data labeling for boxing matches

[francislabountyjr]

Labeling instructions are located here: #2