1. How profitable is developing poker software?
This can be a very good business, but only if you approach it as a long-term project and not as an opportunity to make a quick buck.
Users are willing to pay for tools that improve their gameplay. But our audience is relatively small, customer expectations are extremely high, and the product must be accurate, fast, and useful—otherwise, players will quickly churn.
Working simultaneously on DeepSolver and PLO Genius gives us a broader view of the market. Hold'em and Omaha are different ecosystems, with different training methods and technical challenges, so the source of profit isn't selling the "solver" itself. Profit comes from creating products that players return to again and again because they save time, improve decision-making, and fit into their learning process.
So yes, poker software can be profitable, but it's not easy money. It requires cutting-edge technology, constant product development, robust infrastructure, customer support, and deep subject matter expertise. In this business, profit margins are earned through trust and stability, not hype.
2. Can AI kill the solver market?
AI won't kill the solver market. It will kill bad solver products. LLM can explain concepts, generate hand summaries, or help players improve their learning. But poker is still a mathematical game with extremely precise decision trees. Under the hood, reliable calculations based on game theory are still needed. AI without a solver foundation can confidently spout complete nonsense.
What AI will really change is the interface. In the future, players won't want to manually click through hundreds of nodes. They'll simply ask, "Why does this hand keep betting on the turn?" or "What's the strategic difference between these two flops?" AI can make solver output easier to understand.
So I see AI as a layer on top of solvers, not as a replacement for them.
Read Read3. Do you plan to add support for bomb pots or other new formats?
Yes, of course – new formats are on our radar, including bomb pots.
Since we're developing both DeepSolver and PLO Genius, we're seeing demand from two slightly different worlds – the more structured Hold'em environment and the much more dynamic Omaha ecosystem, where formats like bomb pots are generating a lot of interest, especially in live and private games.
At the same time, we don't want to treat new formats as mere marketing checkboxes. We want to support them properly—with precise calculations and practical training tools; we want to create a user experience that truly helps players improve. For us, it's important not only to calculate the calculations but also to present them in an accessible, understandable, and useful way.
In short, yes, we are interested in adding new formats, but we want to meet the standard that players expect from both DeepSolver and PLO Genius.
4. What will happen to solvers in 5 years?
In five years, solvers will look less like calculators and more like personal trainers.
The first generation of solvers answered the question: "Where is the equilibrium?" The next generation will answer the questions: "How do I even learn this? Where are my leaks? What simplifications can I use in the game? What should I study next?"
The solver market will evolve in three directions:
- Faster calculations
- Clearer explanations
- Greater personalization
Players will spend less time waiting for the simulation to finish and more time practicing and developing game plans.
The products that will survive will not be those with the most complex interfaces, but those that transform complex strategies into simple and easy-to-remember solutions.
5. How do you check the accuracy of calculations within your software?
First, we monitor the technical quality of the calculation itself.
Second, we conduct regression testing on benchmark spots. This way, when improving the engine, we can ensure that the changes are real.
Third, we check the results from a poker perspective, because even a technically correct calculation can be misleading if the ranges, rake model, or tree design are unrealistic.
So for us, validation happens on several levels: the math has to be sound, the results have to be repeatable, and the strategy has to make sense in a real poker context.
6. What is the most difficult thing about developing solvers/simulators?
Turn the solver into a product that can actually be used.
Poker trees are huge even in Hold'em, and in PLO, the combinatorial space and equity interactions become much more complex. But even after you've calculated the spot, problems arise.
How can we transform a huge amount of data into a comprehensible training model? What can be simplified, and how can we help players develop their intuition rather than simply memorize numbers?
7. What percentage of people who buy your software end up never using it?
We don't have such situations. Poker software is a fairly conscious purchase. If someone signs up for DeepSolver or PLO Genius, it usually means they already understand why they need it. For us, the main questions are: how long does the client stay with us, and how often do they return? Does the product remain part of their weekly or monthly learning process?
This is the metric that really matters.
8. Do you play poker? Do your employees play?
Yes, many of us come from a poker background, and many of us still play or study the game.
Some people on the team are former regulars, and this helps the product stay relevant.
At the same time, not every engineer needs to be a high-stakes player. The best product is created when developers build the engine and interface, and poker experts ensure we address current issues.
9. Do your employees who play poker and use the software win?
Some do. But I wouldn't say the solver's quality depends on whether its employees are active, winning players.
The more important question is: do we have people who understand how winning players learn, how they think, and what challenges they face? The answer is, yes, we do have such people.
Poker is changing. Someone who was crushing it five years ago but stopped learning might not be beating the game today. That's why we try to stay close to active professionals, coaches, and serious players. They're the best source of feedback because they immediately notice when something is useful and when it's just theory.
10. What's the most useless feature that users keep asking for?
I wouldn't call our users' requests useless, because they usually stem from a genuine desire to learn something. But if I had to name the least valuable type of request, it would probably be endless customization for very rare edge cases.
People often ask for more filters, more branches, more niche scenarios, and more ways to slice and dice data. On paper, this sounds great. In practice, this often makes the software heavier, the learning curve slower, and the core concepts less clear.
At DeepSolver and PLO Genius, our goal isn't to add complexity for the sake of complexity, but to help players learn faster and make better decisions. Sometimes the best product solution isn't to add another feature, but to remove unnecessary ones.
11. What is missing from your software?
A lot—and that's a good thing. A quality product is never completely finished.
The biggest area for improvement is learning navigation. In DeepSolver and PLO Genius, we already offer powerful calculations, simulators, reports, and visualizations, but the next step is to help users answer more practical questions. What should I learn today? Which leak is costing me the most? What next step will be most effective for my game?
There's still room for usability improvements, new formats, and expanded collaboration and personalization features. PLO Genius already supports mobile devices through responsive web design, but mobile learning can become faster, cleaner, and more intuitive over time.
So yes, there's still a lot missing, and that's encouraging. If we ever felt like we had nothing left to improve, it would probably mean we'd stopped listening to our users.
12. Can you name one feature available in DeepSolver or PLO Genius that competing products don't have?
For me, the main advantage of PLO Genius isn't just the hand import feature, but how it's integrated into the overall learning process.
Players can upload hand histories as .txt and .xml files, or simply as text. These are then compared to a suitable pool of simulations, allowing players to analyze EV loss, decision quality, hand categories, and leaks by position in one place.
This makes the product more than just a solution library—it becomes a practical system for turning real hands into targeted learning.
13. How much time is left in poker?
Poker has a lot more time in store than the pessimists think.
The death of poker has been predicted many times – after the advent of tracking software, HUDs, solvers, funds, and AI. But poker survives because it's not just a mathematical problem. It's also about competition, psychology, bankroll pressure, ego, table selection, live reads, and much more.
Some aspects of poker will become more complex. But poker as a game will remain attractive as long as operators protect recreational players, combat cheating, create fun new formats, and maintain a healthy ecosystem.
Poker won't die because people are learning. It will only die if the game stops inspiring confidence or bringing enjoyment.
Read Read14. Are solvers killing poker?
No. It's not solvers that are killing poker, it's cheating.
A solver is a learning tool, like a chess engine. Problems arise when someone tries to turn a learning tool into real-time hints during a game. That's why fair play has always been our priority.
As far as we know, DeepSolver was the first solver company to publicly introduce the Fair Play Check feature, allowing players and poker rooms to check whether a particular board had been recently solved.
Solvers raise the standard of learning, force clearer thinking, and add depth to the game. The industry's job is to guard the line between legitimate learning and cheating.