poker scientist
7-Year Founder Journey — From Concept to 500 Paying Users
Challenge
Professional poker players knew optimal strategies existed mathematically but couldn't access them in practice. Calculating Nash equilibria for poker requires processing near-infinite game states—every bet size creates exponential branches in already complex decision trees. The computational complexity made brute-force approaches completely impractical. Existing solutions were academic toys that couldn't handle real-world scenarios, leaving serious players with no practical tools.
Solution
We pioneered the first user-facing GTO platform by solving two problems simultaneously: computational efficiency and cognitive accessibility. Built custom abstraction algorithms that reduced infinite situations to tractable calculations, then designed clustering visualizations that revealed strategic patterns humans could actually learn and apply.
Led frontend architecture migration from Angular to Next.js, prioritizing long-term stability and developer velocity. Engineered card matrix UI displaying equity, monetary values, and probabilities in scannable formats. Integrated C++ Nash solver with Node.js API layer, optimizing for both calculation speed and user experience. Designed adaptive visualization components that maintained clarity while presenting complex multi-dimensional data.
7-Year Founder Journey — From Technical Experiment to Profitable SaaS
Poker Scientist is a 7-year founder journey that transformed a computational challenge into a profitable SaaS business with 500 paying users. As Co-founder and Technical Lead, I built what the poker industry considered impossible: making Nash equilibrium strategies accessible and actionable for human players.
The Problem: Why Professional Poker Players Couldn't Access Optimal Strategies
Professional poker players knew optimal strategies existed mathematically but couldn't access them in practice. Calculating Nash equilibria for poker requires processing near-infinite game states—every bet size creates exponential branches in already complex decision trees. The computational complexity made brute-force approaches completely impractical. Existing solutions were academic toys that couldn't handle real-world scenarios, leaving serious players with no practical tools for strategic improvement.
The Solution: Making the Computationally Impossible Become Practically Useful
We pioneered the first user-facing GTO platform by solving two problems simultaneously:
1. Computational Efficiency - Custom abstraction algorithms reduced infinite situations to tractable calculations that could run in real-time 2. Cognitive Accessibility - Clustering visualizations revealed strategic patterns humans could actually learn and apply at the tables
This wasn't just a technical challenge—it required understanding how poker players think, what they need to see, and how to present complex data in ways that improve real-game decision-making.
Founder Journey: Concept to 500 Paying Users
Over 7 years (2018-2025), I led Poker Scientist from initial concept to sustainable business:
- Co-founder & Technical Lead - Not just development, but full business responsibility from product strategy to technical architecture
- Product-Market Fit - Achieved in a highly specialized vertical requiring both technical sophistication and deep domain expertise
- Profitable Operations - Operated profitably for 7 years in a competitive market with three-tier SaaS pricing ($29-$75/month)
- User Impact - 500 paying users who applied learned strategies to improve their real-game win rates
Technical Leadership Highlights
As Technical Lead and Frontend Architect, I made decisions that balanced user value with long-term viability:
- Angular → Next.js Migration - Leadership decision to migrate frontend architecture, improving performance 30%+ while reducing technical debt
- C++ Integration - Cross-language system thinking, integrating custom C++ Nash solver with Node.js API layer for optimal computational speed
- Data Visualization - Designed card matrix UI components displaying equity distributions, monetary values, and probabilities in scannable formats
- Full-Stack Architecture - From payment infrastructure (Stripe & Chargebee) to database design (PostgreSQL for millions of game states)
Why This Matters Beyond Just Code
Poker Scientist proves execution capability over 7 years: not just building features, but sustaining a business, making architectural decisions with long-term consequences, and solving user problems that required both technical depth and domain understanding. The 500 paying users aren't just a metric—they represent validation that we built something people valued enough to pay for, month after month, for 7 years.
Technologies & Skills Demonstrated: Full-stack development, SaaS architecture, payment integration, frontend framework migration, cross-language system design (C++/Node.js), data visualization, API design, PostgreSQL database management, production deployment, technical leadership, product-market fit validation, founder journey
Timeline: 2018-2025 | Role: Co-Founder & Technical Lead
Screenshots
Frontend
Backend
Tools & Services
Database
Impact
Scaled to 500 paying users with three-tier SaaS model ($29-$75/month). Achieved product-market fit in specialized vertical requiring both technical sophistication and domain expertise. Operated profitably for 7 years (2018-2025) in competitive market. Users reported applying learned strategies to improve win rates in real games.
Key Learnings
- Founder perspective: Choosing specialized vertical over broad market—poker players paid $29-75/month for 7 years because we solved THEIR specific problem, not a generic one
- Product-market fit: Achieved sustainable business by focusing on deep domain expertise over broad appeal—specialized verticals can support profitable SaaS when you solve real pain points
- Technical leadership: Angular → Next.js migration decision improved performance 30%+ while reducing technical debt—sometimes you need to make hard architectural choices for long-term viability