Analytics & model workflows
Look beyond the output. Evaluate how data is collected, models are validated and findings reach the people making hockey decisions.
Data quality · Evaluation · Decision workflows
Independent hockey consulting
An outside perspective. An inside understanding.
Independent analytics, AI and technology consulting for NHL organizations—connecting the realities of hockey with the systems behind better-informed decisions.
Built for NHL organizations. Independent of any club or league.
01 / The work
A new tool is not always the answer. Start with the decision, understand the workflow, then determine what will actually help.
Look beyond the output. Evaluate how data is collected, models are validated and findings reach the people making hockey decisions.
Data quality · Evaluation · Decision workflows
Separate a useful application from a compelling demo. Define bounded use cases, test limitations and keep human review and sensitive-data boundaries in the design.
Use-case discovery · Guardrails · Scoped pilots
Build around the way your staff actually works. Connect scouting, coaching and analysis with focused tools that make information easier to find, understand and use.
Workflow design · Prototypes · Integrations
Add an independent technical perspective without creating another silo. Pressure-test vendor claims, clarify build-versus-buy choices and shape a practical roadmap.
Vendor diligence · Architecture · Technical direction
02 / The bridge
The strongest system connects the people already in the room.
A scout sees context. An analyst tests a pattern. A coach needs something usable. Technical work matters when it helps those perspectives meet—without flattening the differences that make each valuable.
Name the decision and who needs to make it before choosing a platform, a model or a metric.
Expose assumptions, data gaps and failure modes. A clear limitation is more useful than false precision.
Keep feedback close, ownership clear and tools understandable enough to question—not just operate.
03 / The background
Donny Grover brings a professional playing background to his work as a technology founder and fractional CTO. The connection is practical: understand the environment, ask better questions and make the work useful.
The technology background2001–2005 / NCAA Division I
Four seasons of NCAA Division I hockey at Northeastern, competing in Hockey East.
Professional hockey / North America
AHL experience with the Lowell Lock Monsters and Hershey Bears, as well as professional play in the ECHL.
Professional hockey / Europe
A playing career that also included professional hockey in Germany and Denmark.
Technology / Founder & fractional CTO
Founder of Grover Web Design and fractional CTO. His public professional bio lists electrical engineering at Northeastern with a business minor.
04 / Ways to work together
A clear question. An agreed scope. Something useful to take forward. Choose the level of support that fits the work, not the other way around.
01 / Understand
Start with one decision, workflow or technical question. Establish what exists, what is uncertain and what deserves a closer look.
02 / Test
Explore a specific use case before committing to a larger rollout. Agree on the boundaries and what a useful result would look like.
03 / Support
Keep an independent technical sounding board close to the work. Support your staff as questions, priorities and vendor decisions evolve.
Deliverables, timing and fees are agreed for each engagement. No promised wins, rankings or competitive outcomes.
05 / Field notes
The useful integration is not another screen: it is a coaching question connected to traceable events, reviewable clips, and an accountable owner. Define source IDs, correction…
An AI roadmap should identify what can be tested safely, what must stay protected, and who can stop a pilot. Use this practical framework for vendor…
A useful audit traces a hockey metric from its source to the decision it is meant to support. Here is how to test context, reproducibility, out-of-sample…
06 / A few useful answers
No. This is independent, third-party consulting designed for NHL organizations. No club or league affiliation, endorsement or existing client relationship is implied.
Yes. The intent is to complement the people already doing the work: add capacity, offer an outside review or help connect departments around a specific question. Ownership and responsibilities are agreed at the outset.
Start with a defined problem, authorized data and an evaluation plan. Account for inaccurate outputs, privacy, access controls and vendor data use. Keep consequential decisions with accountable people, and be willing to conclude that AI is not the right tool.
No. An initial conversation can stay at the level of goals, constraints and workflows. Any later access should be explicitly agreed, limited to what is necessary and handled through approved systems—not a public contact form.
No. The goal is to make evidence more usable and limitations more visible, not to outsource hockey judgment to a dashboard or model. Scouting, coaching and analysis bring different context; the work should connect them.
Begin with a conversation about a real question. From there, agree on a focused diagnostic or a bounded pilot, with written scope, deliverables and evaluation criteria. Timing and fees depend on that scope; no competitive results are promised.
Start with a real question
A workflow that is not connecting. A vendor claim worth testing. An AI idea that needs a grounded second look. Let’s talk about the work.
Keep the first message high-level. Please do not send confidential player, medical or club information through the public contact form.