Regulated platforms • Revenue-critical systems • Stability-first leadership • Advanced AI modelling • Innovation

Case studies

Problem → approach → measurable impact.

NFL simulation & pricing platform

Real-time • Production
Problem Broadcaster needed higher-quality, telemetry-driven odds and derived markets from live NFL data feeds.
What I did Built AI powered models based on 20+ years of previous real NFL data. This powered prediction engines, APIs, pipelines, and cloud-native services with explicit SLOs (Service Level Objective), observability, and cost controls.
Impact Launched live in-play markets and probabilities on ESPN via Adrenaline’s integration during the 2025 NFL season, producing clear revenue streams and obvious opportunities to expand into Live Sports Betting.
AWSTypeScriptPython AI / ML toolingLive data

Virtual American football

Simulation • Data • Betting platform
Problem The client needed a realistic virtual American football sports game, but lacked the knowledge and underlying data to build it.
What I did Built an AI powered simulation platform using real NFL and college football data, running 2M+ simulated games to generate realistic outcomes and telemetry using 20+ years of historical data and AI model based prediction, packaged for straightforward integration into the client’s product. This product generated 50+ betting markets based on these simulations, creating a rich experience for the user.
Impact Built a virtual American football game scheduled for US casino launch in 2026, proving the company’s predictive models were strong enough to support a move into live events data.
Example virtual football game screenshot
Simulation AI / ML tooling Data engineering 2M+ games Integration US casinos

Regulated wagering ops maturity uplift

Ops • Governance
Problem Production incidents + unclear ownership. Audit expectations rising. Delivery slowed by ambiguity.
What I did Introduced RACI (Responsible Accountable Consulted Informed), incident governance, on-call tiers, escalation, and exec-facing risk reporting.
Impact ↓ incident MTTR (Mean Time To Resolution) by 90% • ↑ delivery predictability • Gave execs predictable delivery and audit-ready operational reporting.
Incident MgmtSupport modelRisk reporting

Embedded betting launch (UK)

Delivery • Client alignment
Problem Pre-launch issues threatened go-live of a new embedded sportsbook product.
What I did Took go/no-go ownership with client, froze scope, re-cut the critical path, and reset launch criteria with exec sponsors.
Impact Unblocked a strategic product launch, enabling UK market entry and unlocking new revenue streams, while repositioning the business toward an out-of-the-box sportsbook platform offering.
Example Screenshot of racing app
Program delivery Stakeholder alignment Go-live readiness Regulated market

Smarter pricing, higher margins

Real-time • Betting
Problem Operators needed to optimise take (hold) without degrading the customer experience. Traditional pricing applied a largely static vig across markets, ignoring real bettor behaviour—leaving money on the table in some markets and reducing competitiveness in others.
What I did Designed and delivered an advanced AI-based pricing platform using real-time behavioural and market data to predict bettor preferences and dynamically optimise vig, integrating via APIs into live, risk-controlled trading workflows.
Impact In controlled trials, the system delivered material improvements in operator take versus a typical 10% hold.
Example pricing dashboard screenshot
Odds Markets AI / ML tooling Improved Profit

Restoring cloud cost predictability at scale

Costs • Hosting
Problem High cost of AWS hosting.
What I did Reviewed AWS usage, identified cost-saving opportunities, and implemented optimisations that reduced hosting costs by 50%.
Impact ~95% reduction in peak spikes over 18 months (from ~20k to ~1–2k). Consistent 50% reduction in monthly spend over 2 years. Restored cost predictability, removed growth constraint, improved margins.
AWS Cost alignment

Writing

Some examples of topics that I write about, what occupies my mind 24/7.

Operations

Lessons for Engineering Leaders

Things engineering leaders get wrong (and I’ve definitely got wrong at least once) #374

Incident mgmtTech debtGovernance

Get In Touch

If your platform is holding the business back, or starting to worry the board, connect to talk about what “stable and scalable” actually looks like in your context.

For CTO / VP Engineering roles, or similair.

Email me Message on LinkedIn Book 30 mins (Calendly)