All case studies

Case study · Sportsbook · 2025

BetForge

Sportsbook trading desk goes AI-native, without slowing the lines.

A multi-market sportsbook with 400 staff embedded AI across odds compilation, risk, and support — same risk veto, faster cadence.

At a glance

  • 400Headcount
  • 4 weeksEngagement
  • 10Pioneers
  • 6Markets
  • 24/7Operating

The brief

AI inside the trading desk, with risk veto on every line.

See how we did it
Abstract illustration

Cohort

  • C-level (CEO, COO)2
  • Heads of Department5
  • AI pioneers10

Drawn from Trading, Risk, Payments, Customer Support, and BI — the five functions where speed and accuracy compound.

What the pioneers shipped

Bots that think with traders, not for them.

Want this for your team?

Live trading copilot

Watches markets, suggests line moves with rationale. Trader signs every change.

Risk-team alert triage

Surfaces unusual betting patterns, links to similar past cases, drafts the analyst note.

Settlement reconciliation

Cross-check between trading platform and payments rail. Discrepancies flagged before payout.

Customer support classification

Tier-1 questions auto-routed; complex disputes escalated with full context.

In-play exposure summary

Natural-language summaries of in-play volumes and book exposure — refreshed every 60 seconds.

“Faster cadence — same risk veto on every line.”

Approach

Four weeks. Risk team in the room from week one.

01 / 04Week 1

Discover

Mapped trading workflow, risk gates, settlement rails. Risk team and traders co-defined what could be assisted vs. authored.

02 / 04Week 2–3

Build

Pioneers built first cuts inside their function. Every model output passes through a risk-team review before staging.

03 / 04Week 3–4

Validate

Three Friday demos. Each model challenged by traders running their own A/B against the agent's suggestions.

04 / 04Week 4+

Compound

Skills shipped to internal marketplace. Cross-market learning loop instrumented; suspended-market actions stay fully manual.

Impact · Year one

Outcomes leadership signed off on.

  • Annual savings

    €3.2M

    Across the engagement's first twelve months.

  • FTE optimized

    24

    Full-time equivalents freed for higher-leverage work.

  • Business satisfaction

    9.4 / 10

    Post-engagement leadership review.

Why it didn't go off the rails

Trader veto. Risk veto. No exceptions.

Traders sign every line move

Agent proposes; trader signs. No automated price changes on live markets.

Risk team has model veto

Each model has a risk owner. Anything not signed off by risk doesn't ship.

No automation on suspended markets

Suspended markets are always handled manually. Audit trail per intervention.

Cross-jurisdiction logging

Same query in different markets returns the same documented answer — every regulator inspection passes.

Where to start

Book an org assessment.
Ninety minutes, no slides.

We'll spend an hour with your leadership team mapping where AI creates the most leverage in your operation — and what an overlay would look like in practice.