// Brett Ragozzine, PhD · Strong Force Analytics
I design and deploy agentic AI systems, automated workflows, and intelligent pipelines that run without you. PhD physicist. 20 years applied ML. Now building the infrastructure that makes data science disappear into the background — and revenue appear in the foreground.
// What I Build
Three layers. Each one more autonomous than the last. Most clients start with a workflow and graduate to agents.
Multi-step pipelines that pull data, run models, detect patterns, and deliver actionable outputs — on a schedule, without human intervention. Built once, runs forever.
Goal-driven systems that make decisions, use tools, handle exceptions, and escalate only when necessary. They work while you sleep. Powered by LLMs + custom logic.
Orchestrated networks of AI agents collaborating toward a shared goal. Multi-agent coordination, specialized roles, shared memory. This is the frontier — and where the biggest leverage lives.
// Case Studies
Real projects. Real outcomes. The pattern is always the same: find the data nobody is looking at, build the system that acts on it automatically.
Finicity / Mastercard · Fintech
Built and led two data science teams to train LLMs for transaction categorization, income calculation, and entity recognition across multiple countries and languages. Designed a scalable architecture that expanded into Canada, Australia, and Brazil without proportional team growth.
→ Patent issued · Expanded to 3 new markets · Semi-automated labeling system reduced data ops load by 60%+
Purple · E-Commerce
Built a model quantifying exactly how delayed delivery timelines affected sales conversion and customer lifetime value — dollar amounts that were previously invisible to leadership. Also delivered Experian Mosaic targeting models to improve ROAS and introduced Tableau across the organization.
→ Delivered first data-driven ROAS model · Backorder impact quantified in dollars for executive decisions
Wrench.ai · MarTech
Invented and patented models that predicted customer personality from text, then used those profiles to recommend ideal messaging strategy, contact channel, and product affinity per lead. Built churn prediction, lead scoring, and a product recommender — all wired into a single customer intelligence platform.
→ Patent issued · Full customer intelligence stack deployed in production
Grocery Retail Client · Strong Force Analytics
Retailers were losing sales daily to out-of-stock events that nobody saw coming. Built a recommendation system that predicted restocking events throughout the day based on sales velocity, time patterns, and inventory data — surfacing what the data already knew before shelves emptied.
→ Measurable same-day sales lift · Reduced manual restocking guesswork
// Technical Stack
20 years of applied ML across fintech, retail, trading, and astrophysics — now extended into agentic AI systems and autonomous trading software.
◆ = actively building now | gold dots = new capabilities added 2025–2026
// Products
The same leverage I apply for clients, applied to my own products. Automated trading software for NinjaTrader 8, deployed with a custom licensing and subscription system.
A complete ecosystem of automated trading strategies and order-flow analysis tools for NinjaTrader 8. Includes Smart Money Concepts strategies, Opening Range Breakout automation, Effort vs Result order-flow divergence detection, Volume Profile with heatmap, and Footprint charting — all with professional WPF dashboards and Airtable-backed licensing.
Indicators · Strategies · Full Order Management · Live Licensing System
Opening Range Breakout with range-percentage targets, pullback re-entry system, and full risk controls. Built around the open — one of the highest-probability windows in futures trading.
$399 lifetime · $79/mo
Effort vs Result — trades delta divergence and continuation signals with 7 calculation modes and a modular filter system. Order-flow intelligence, automated.
$499 lifetime · $79/mo
// About
I started by modeling dark matter in galaxy clusters using weak gravitational lensing — writing image processing pipelines to analyze data from the Hubble Space Telescope, speeding up established scientific software by 450–5000%, and earning a PhD in Physics and Astronomy from Ohio University.
That trained something useful: the ability to find signal in noise at massive scale, formalize problems mathematically, and build systems that actually run in production — not just in notebooks.
I took that into data science — Zions Bank, InsideSales, Purple, CompuCom, Wrench.ai — and eventually to Director of Data Science at Finicity/Mastercard, where I hired and led two teams, earned two patents, and expanded ML systems into international markets.
Now I run Strong Force Analytics with a development team in Bangladesh, building AI workflows, agentic systems, and automated trading products. The goal isn't to do more data science. It's to build systems that do it — while I scale.
Whether you need a workflow that runs without you, an agent that makes decisions, or a complete agentic system — let's talk about what's possible.
Start the Conversation