The Hockey Brain

AI agent organizations · hockey intelligence

The Hockey Brain

Stop renting analytics. Start running a machine.

Outwork the biggest data departments. One human lead, a full AI agent organization on our platform—smarter beats bigger.

Lean team. Massive output. You bring the judgment—we help you deploy the machine.

Demo Club · PlatformLIVEAnalytics DirectorHockey OperationsTranslates goals into agent tasksAI AGENT TEAMData LeadData & PipelineClaude · AI agentRaw ingestion to modelsAnalytics LeadInsights & ModelsClaude · AI agentQuality of all outputDelivery LeadReports & OutputClaude · AI agentClient-facing deliveryData EngineerAI · agentNormalises game dataStats AnalystAI · agentAdvanced metricsPerformanceAI · agentPer-player reportsScoutingAI · agentDraft analysisDashboardAI · agentReal-time analyticsReport AgentAI · agentAuto report generationDEMO CLUB · STAFF MAP
Same game · unfair advantage

Stack roles like the big clubs— without their headcount.

In the platform you wire a full tree: leadership, leads, specialists—agents that own pipeline, insights, delivery and client-facing output. The sketch below is simplified; the real app goes as deep as you need. One person can run the control room while agents do the volume work—so you compete on architecture and execution, not office size.

The highlighted node is where governance usually sits: strategy stays human; everything underneath can scale with agents. Demo labels use a generic demo club—your titles, your rules.

Connects to your stack

League APIsVideoTrackingSheetsBI tools

Demo · Staff map

General Manager

Club leadership

Head Coach

Bench & systems

Active

Director, Hockey Analytics

Models & reporting

Chief Amateur Scout

Draft & prospects

AI agents under analytics

Scouting agent

AI · tape & prospects

Reporting agent

AI · models & alerts

The product·demo club · illustrative UI

Your AI org—not a slide. A system.

Real screens from the platform: full agent hierarchy first—then the rest of the workspace. This is what the new era looks like when you stop renting analytics by the hour and start owning an organization of agents.

The Hockey Brain platform — full AI agent organization chart (demo club)
The Hockey Brain app — demo club, analytics view
The Hockey Brain app — demo club, detail view

AI agent org chart

Manage your goals

Goals that actually ship.

One mission, visible metrics, and agent-assisted workflows that trace back to what you're trying to prove on the ice—without a room of analysts babysitting spreadsheets. Below: how serious programs structure what they measure.

  1. 01

    Define the season objective

    Set the north star—playoff seeding, development minutes, cap efficiency, special teams rank. Everyone sees the same headline goal.

  2. 02

    Translate into analytics targets

    Break the objective into measurable signals: rush xG share, slot shots against, prospect NHLe, goalie workload. Targets roll up to the club goal.

  3. 03

    Review, prioritize, act

    Humans approve direction; agents surface options and keep metrics fresh. Everyone stays aligned to what you said mattered in step one—without a big analytics staff doing manual busywork.

What clubs typically track

Process & finishing

  • 5v5 xGF / 60
  • Slot + rush chances
  • Shooting % vs expected

Special teams & game state

  • PP shot rate & entries
  • PK slot suppression
  • Score-adjusted shares

Roster & pathway

  • Prospect cohort curves
  • Contract surplus value
  • Injury / load monitoring

Opposition intel

  • Line matchup history
  • Goalie tendencies
  • Bench usage patterns

Labels are illustrative; your league and data vendors define the exact fields. We help you pick a coherent set—often with AI agents keeping dashboards and watchlists current—so coaches, scouts and management are not optimizing against different scoreboards.

Why it works

Smarter beats bigger. Every time.

The clubs that win don't have the biggest budgets—they have the best systems. An AI agent organization doesn't need more headcount. It needs sharper architecture—and the nerve to use it.

1operator

runs the full agent stack—from setting goals to overseeing output to presenting to ownership. No team required.

24/7coverage

Agents handle scouting prep, opponent intel, prospect monitoring, and cap modeling continuously—not when your analyst has bandwidth.

2business days

to first response on new engagements. No sales cycle. Scope it, start it.

Shooting Percentage Evolution

Last 3 seasons + predictive model

Shooting Percentage Evolution Over Time8%9%10%11%12%13%14%S1S2S3PredActualLeague AvgPrediction

Advanced modeling reveals upward trajectory exceeding league baseline, with predictive confidence intervals guiding tactical roster decisions.

How we work

Advisory or platform. Or both.

Start with a focused sprint. Graduate to the full agent stack when you want the game to run 24/7.

Advisory

Human judgment, on demand

Senior analytics applied where it matters: framing the right questions, interpreting model output, presenting findings to ownership and coaching staff. The parts agents don't own.

  • Sprint engagements (4–8 weeks)
  • Strategy & framing
  • Stakeholder communication
  • Project-based reporting
Book a consultation
Platform

Platform

Your AI agent organization

A full hierarchy of configured AI agents—leads, specialists, delivery—running continuously. Scouting, opponent prep, prospect monitoring, cap modeling. You own the org chart.

  • Monthly subscription
  • Agent org chart in the platform
  • 24/7 continuous coverage
  • Scale by adding agents
See the platform
What agents handle

Six capabilities. One machine.

Each capability maps to an agent type in the platform. Turn on what matters—mute the rest.

For hockey clubs & organizations

Built for clubs that refuse to lose on process.

We work with pro clubs, junior organizations, national teams, and player agencies that want data depth without building an internal department. Senior-level analytics from Sweden—people who understand both the code and the game.

  • Drafts, free agency, trade deadlines, and roster planning.
  • Scouting models that fit your existing video and data stack.
  • Processes coaches, scouts, and management actually use.

How a typical engagement starts

  1. 130–45 min call: your league, data sources, and staff structure.
  2. 2We propose 1–2 concrete projects that move the needle in 4–8 weeks.
  3. 3Agree on scope and format — then start delivering.

Remote from Stockholm. Time-zone agnostic. Answers within two business days.

Book an intro call
Get started

Stop renting analytics.
Start running a machine.

Book a 30-minute intro call. We'll scope an engagement or pilot that fits your league, data, and timeline. No long sales cycle. Two business days to first response.