AI Tracking Gets New Fuel: What Product Teams Can Learn from SkillCorner's $60M Raise
- SkillCorner secures $60M from Silversmith Capital Partners to accelerate multi-sport growth in North America
- AI tracking now spans 180+ competitions and serves 300+ organizations globally
- NBA remains in the ownership group alongside Silversmith and the founding team
- Founded in 2016, the company has grown past 100 employees and is expanding beyond soccer into basketball and American football
- Funding targets data engineering, product development, and new decision-ready metrics for pro and collegiate teams
Why this matters for product leaders
SkillCorner's software-first tracking model removes hardware friction by extracting data from standard broadcasts. That choice speeds distribution, simplifies onboarding, and lowers cost to serve. It's a blueprint: pick an input everyone already has, then build the modeling and metrics layer that turns it into decisions.
The fresh capital signals a push from single-sport depth to multi-sport breadth. Product teams can read this as a roadmap shift: expand addressable market via model generalization, then compound value with sport-specific metrics the market actually uses.
How the tech scales
Single-camera automated tracking estimates player and ball positions directly from broadcast feeds. The system infers off-camera locations to produce complete sequences without stadium-installed rigs. This reduces deployment friction and enables rapid coverage across leagues and levels.
Under the hood, scale depends on data labeling pipelines, model retraining cadence, and rigorous QA. Accuracy on messy, real-world footage is the constraint; product velocity comes from tooling that shortens the feedback loop between model updates and metric releases.
North America expansion: multi-sport, multi-tier
The company is prioritizing basketball and American football across professional and collegiate levels. That opens the door to standardized data products used from recruitment to match prep, with consistent APIs across sports.
For product managers, the practical move is to design modular schemas now. Keep core entities (players, possessions, events) consistent, then layer sport-specific taxonomies for plays, roles, and context.
Data engineering priorities funded by the raise
- Foundation data: scalable ingestion from broadcast sources, robust event synchronization, and deduplication
- Model generalization: cross-sport feature libraries, transfer learning strategy, and evaluation sets by camera format
- Metric R&D: recruitment and tactics metrics validated against outcomes (win contribution, injury risk proxies, development curves)
- Latency and reliability: near real-time inference, confidence scoring, and service-level guarantees for teams and broadcasters
- Tooling: reviewer workflows, synthetic data for edge cases, and automated regression checks before each model push
Metrics that drive decisions
Teams care less about raw tracking and more about actionable metrics. Expect emphasis on off-ball value, spatial pressure, spacing integrity, fatigue indicators, and play similarity indices. The winners will be the metrics that correlate with recruitment hits and tactical outcomes, not just nice visuals.
Build product stories around specific jobs-to-be-done: shortlist prospects across leagues, benchmark player development year over year, stress-test matchups, and simulate tactical adjustments.
Strategic implications for collegiate and youth
As collegiate programs adopt broadcast-based tracking, demand will pull downstream. Without the need for stadium hardware, smaller programs can access pro-grade analytics. That widens the data funnel and improves long-term player evaluation models.
For product teams, plan for identity resolution across fragmented data sources and shifting rosters. Invest early in permissions, consent management, and data rights contracts with conferences and broadcasters.
Build vs. buy guidance for teams and leagues
- Buy if you need speed, breadth of coverage, and reliable benchmarks across competitions
- Build if you have proprietary footage rights and a unique model or workflow advantage you can maintain
- Hybrid works: buy core tracking, build custom layers for scouting, development, and playbook analysis
Risks to manage
- Model drift across venues, camera angles, and broadcast standards
- Bias in training data that skews player evaluations by league or style
- Rights and licensing complexity with broadcasters and conferences
- Privacy and consent for youth and collegiate athletes
- Vendor lock-in: export guarantees and schema stability matter
What to watch next
- North American partnerships across major leagues and the NCAA tier
- API products and SDKs that let teams blend tracking with internal scouting and medical data
- Transparent accuracy benchmarks versus optical systems
- Pricing aligned to outcomes (recruitment success, win contribution) rather than raw data volume
Ownership now includes Silversmith Capital Partners and the NBA. For context on the stakeholders, see NBA and Silversmith Capital Partners. The company continues to operate independently from Paris with a team of 100+ employees.
Quick checklist for product teams
- Define the minimal metric set that changes decisions this season
- Model evaluation protocol by sport, competition, and venue format
- Data contracts: rights, retention, consent, and audit trails
- Roadmap split: platform reliability first, sport-specific features second
- Feedback loop: embed analysts and coaches in discovery and QA
If you're upskilling your team on AI product foundations and data strategy, explore role-based resources at Complete AI Training.
Your membership also unlocks:
