From the College Athletics edition of September 21, 2026
USC Viterbi data mart nears operational use with football’s director of AI
University of Southern California / Daily Trojan · College athletics / football coaching analytics data engineering · United States (USC Athletics, Los Angeles)
- Publisher
- Daily Trojan
- Original publication
- September 16, 2026
- Source retrieved
- 2026-09-22
What happened
A 12-person Viterbi data-engineering team has spent about eight months building a football statistics data mart with USC football so analysts can work from analysis-ready play history rather than inconsistent raw feeds. The team coordinates with Conor McQuiston, director of artificial intelligence for USC football. Phase 1 (every play for every team over ~10 years) is described as about to go operational for the current season; phase 2 adds per-player role detail. Staff explicitly defer fourth-down predictive products until data quality is ready. No measured time-savings or competitive outcomes are reported.
Why it matters
Newly relevant Power conference implementation signal not in the archive and not in the 2026-09-20 edition: an on-staff AI director plus academic engineering partnership building the data foundation before flashy models—complements archived ASU/Intel on-device play-card automation with a different architecture (central data mart vs local agent).
Evidence and measured results
Daily Trojan feature (2026-09-16) quoting Wilcox and Prabhu on operational timing, McQuiston coordination, phase scope, and deferral of predictive decision tools; reference to 2025 Science and Medicine in Football decision-bottleneck study.
Limitations and uncertainty
Campus newspaper, not an audit; “about to go operational” is intent; no metrics; AI-director title is reported; privacy posture unspecified.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-22; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Football analysts spend scarce week-of-game hours cleaning inconsistent multi-vendor play data instead of producing coaching insights.
- Stakeholders
- Director of football AI/analytics
- Head coach / coordinators
- Athletics CIO/IT
- Academic engineering partner
- Compliance (data use)
- Discovery questions
- Who owns the football analytics stack and data contracts today?
- Which play/film vendors feed the staff, and how messy are joins?
- Is there an AI/analytics director role with authority over tools?
- Which coach questions must wait until data quality is fixed?
- Value hypothesis
- If a curated multi-year play mart reaches operational use, analyst time can shift from wrangling to insight generation—without claiming automatic win-rate lift.
- Potential engagement
- Four-to-six-week discovery of current football data sources, a mart MVP scope, and access matrix with the AI/analytics lead.
- Unsupported claims
- Guaranteed prep-time percentage savings
- Ready fourth-down or play-calling automation
- That USC’s model generalizes to every Power program’s vendor mix
Pre-sales engineering
Role takeaway
- Fit
- Athletics–engineering data platform for football play/player history feeding human analysts and later models.
- Architecture and integration
- Batch ingest of historical play feeds into a governed mart; semantic layer for analysts; separate sandbox for AI experiments consuming curated tables only.
- Prerequisites
- Multi-year play-level source access
- Named AI/analytics product owner
- Identity and environment separation for athletics IP
- Constraints
- Predictive decision products explicitly deferred in source
- Campus partnership staffing may be seasonal
- No published performance SLOs yet
- Security
- Restrict opponent/scheme data; no uncontrolled export of proprietary tendencies; separate student-athlete biometric stores if ever added.
- Proof of value
- Analyst hours on wrangling vs insight before/after phase 1
- Documented coach questions newly answerable
- Data-quality defect rate on play keys
Delivery
Role takeaway
- Work
- Phase-1 mart go-live, analyst training, feedback cadence with football AI director, phase-2 player-role enrichment backlog.
- Dependencies
- Vendor data licenses
- Athletics–academic collaboration agreement
- Identity/access provisioning
- Ownership
- Football AI/analytics director owns use cases; data engineering owns pipelines and quality.
- Skills and adoption
- Analysts adopt mart schemas; coaches learn what is queryable vs still manual.
- Governance checkpoints
- Written phase-1 acceptance criteria
- Ban on unvalidated auto-recommendation tools
- Quarterly access review
- Acceptance criteria
- Phase-1 play history queryable by named analysts
- Documented data dictionary
- No unauthorized external data shares in pilot
- Risks
- Operational claim slips without quality gates
- Model demos before data contracts mature
- IP leakage via academic collaboration tooling
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Prioritize play/player event schemas, quality checks, and analyst-facing tables before coaching decision models; keep model experiments in the AI director’s sandbox consuming curated marts rather than raw vendor dumps.
Governance
Who approves, reviews and stays accountable for outcomes?
Name ownership between athletics analytics and academic engineering partners; define which coaches may query which grain of opponent/player data; document that predictive play-calling tools are out of scope until validated.
Security and privacy
What data, permissions and controls need testing?
Opponent and proprietary scheme data are competitive secrets—prefer controlled campus identity access; student-athlete biometric streams are not described here and should not be silently added to the same mart without consent review.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Graduate engineering collaborators and football analysts need shared documentation so knowledge does not leave with a single student lead.
Procurement
What should contracts, pricing and exit terms secure?
Fund data-platform staffing and source licenses before buying another black-box play-calling AI.
Operating model
Which teams own the service once it runs?
Keep human AI/analytics ownership between mart and coaches; treat academic partners as pipeline contributors, not game-day decision makers.
Publication history
- 2026-09-21College Athletics · Issue 154 resources
Stable resource ID: usc-viterbi-football-datamart-2026