Gameer
Turn an idea into a game, and a game into a business
Gameer turns a prompt into a game people can play. It is becoming a place where anyone can make a world, publish it, find an audience, and earn from what they create.
Data Scientist, Programmer, Founder
Anya Ozmen
Anya Ozmen is an immigrant technical founder. Her mission is to financially and creatively empower consumers and businesses. Her ventures include Redko, Redvest, and Gameer. Together, they help democratize software creation and access to education, turning ideas into products people can launch, own, and earn from.
SF / CA3×Founder
1M+Redvest users
Top 30iOS Finance
Top 200TechCrunch Disrupt
40+Redko clients
Companies
Companies that help people learn, create, launch, and earn.
Gameer
Gameer turns a prompt into a game people can play. It is becoming a place where anyone can make a world, publish it, find an audience, and earn from what they create.
Redvest
A gamified financial education platform where over one million learners across 40+ countries practice investing, build confidence, and learn without risking real money.
Redko
A technology and AI studio that has built products, platforms, and data systems for more than 40 enterprise and startup clients.
Pleasure Lab
Pleasure Lab is an early personal wellness prototype designed to keep sensitive data encrypted on your own device by default. Its sensor data is treated as a limited physiological signal, never proof of desire, pleasure, or consent.
Where the founder story began
My first company explored the intersection of quantitative investing and intelligent trading systems, a seed that later grew into a much bigger mission around financial access.
Philanthropy
Closing the gender gap in health, expanding financial education, and helping more people build. See the philanthropy page
Financial access
01The Redvest Financial Literacy Foundation supports practical financial education. Five percent of proceeds from my consultation calls are donated to the initiative.
Innovator in Residence
02USC Viterbi Startup Garage
As Innovator in Residence at USC Viterbi Startup Garage in Spring 2024, I worked with student founders on product, early traction, and building technology from zero. I also interviewed students for the Summer Smasher cohort.
USC Viterbi Startup GarageMentor
03FoundersBoost LA
Mentor for the FoundersBoost LA Fall 2021 program, giving early founders practical feedback on pitch decks, business models, and finding their first market.
FoundersBoostWomen who build
04Mentoring, speaking, and showing the work so the next generation sees technical entrepreneurship as something they can do.
Women in Tech finalistCauses I care about
I want to help close gaps in how women's physical and mental health are studied, diagnosed, and treated. The goal is better quality of life: symptoms taken seriously, clearer answers, and care people can actually access.
I want to combine brain imaging, mathematical matrices, and AI tools to simulate how biological systems change. One research direction is AI simulation based mental health diagnostics. Another is women's health. The aim is to test hypotheses and work toward better answers, with clinical validation built into the research.
Read the idea and the sciencePeople should be able to examine the quality of the evidence behind a claim. MedLit helps readers assess a paper's methods, risk of bias, and reporting, with its judgments linked to passages in the source.
Explore MedLitSynaptik explores research informed practice games for attention, emotional regulation, and everyday mental wellbeing. It is a place to try these ideas through interaction, with clear limits on what the games can measure.
Explore SynaptikProjects for good
Appraises scientific papers with study-design-specific questions, supporting source passages, and explicit uncertainty. Its score describes appraisal traceability, not whether a paper is true. Connect your own AI provider, or explore the example without an account.
Short practice games for symptoms people actually live with: attention, impulse control, rumination, anxiety, and mood. Each game follows a named protocol, records only what it can honestly measure, and never punishes you for stopping.
Hackathon in the making
The Virtual Lab Hackathon is taking shape in San Francisco. I am inviting organizations, universities, and labs to help bring researchers and builders together, contribute resources, and turn ambitious ideas into working prototypes.
Bring funding, tools, expertise, or a challenge worth solving. Let's shape a role that makes sense for your organization and the teams participating.
Previous research, mentorship, and speaking. Hackathon partnerships will be announced separately.
For hackathon judging, panels, mentorship programs, women's health projects, or any philanthropy collaboration, send a note with the date and what you have in mind.
Journey
Creating products and memories.





100 Most Influential Turks in USA, one of four Under 30 honorees
View sourceEarlier
Press and profiles
More articles and interviews are in the full archive.
Data science and research
Cognitive psychology, predictive modeling, and questions that deserve better tools.
Mathematical matrices, brain imaging, and AI tools for studying mental disorders and women's health. A research direction toward better explanations and, eventually, validated diagnostics.
Explore the research ideaCognitive psychology research on whether newly learned words can influence unconscious semantic priming, and how quickly that processing emerges.
Read the sourceDrivenData · Top 4th percentileBuilt a binary classification model that placed in the top fourth percentile among more than 5,000 competitors.
Read the sourceThe next chapter / Redko Labs
I’m building Redko Labs, an impact AI and deep tech incubator for technical founders, students, and researchers. Bring a research direction, an early company, or the curiosity to find your next big problem.
Co-founder matching, hands-on product support, experienced mentors, and fundraising preparation. Rolling enrollment, with or without an idea.
Selected writing
Machine learning, cognitive science, and future systems.
Biological simulation
Scientific literacy
Artificial intelligence
Cognitive science
Future systems
Open source and impact projects
Public tools built to make useful knowledge and technology easier to reach. The full list is on GitHub.
Structured appraisal of study methods, risk of bias, and reporting, with supporting source passages and explicit uncertainty.
TypeScript, Expo, React Native WebRuns the same task with and without an agent skill file and reports whether the skill changed the outcome, plus its token cost.
JavaScript CLIRegression tests for context compression: run a prompt against full and compressed context and catch answers that silently drift.
JavaScriptWrite one manifest and compile instruction files for Claude Code, Codex, Cursor, Gemini, and Copilot instead of maintaining five.
JavaScriptGenerates App Store and Google Play screenshots from a single prompt, resizes them to every required store size, and uploads them.
JavaScript CLI, npmHow Gameer started
I had stories I wanted people to step into, not just read. I taught myself 3D and started trying to turn those stories into games.
A single scene could take a full day to render. I loved making the world. I hated how much the tools got in the way.
That frustration stayed with me. I created Gameer for people who have ideas but not necessarily the money, team, or technical background to build them. I want them to make something people can enjoy, publish it, and earn from what they create.
The journey
Founder, game creation with AI
Founder, gamified financial education
CEO, AI and product studio
Demand forecasting, inventory optimization, market analysis, and machine learning systems at Advanced Nutrients.
Data analysis, targeted messaging, automation, and product experiments for a consumer app.
Machine learning competitions for blood donation, flu vaccination, and dengue fever prediction.
Coded experiments and studied subconscious processing; presented research through the Psychonomic Society and Western Psychological Association.
A development internship that sharpened how I think about audience, entertainment, and the experience around a story.
A development internship that taught me to think in story and worlds before those instincts met software and games.
First company, quantitative machine learning
Mission
As a young immigrant who moved to the United States by myself, I had to learn all sorts of skills alone. It was a very challenging time, but it taught me to be stronger and inspired me to create products that help people teach themselves, too. I was a rebellious child who did not fit in. I liked learning by doing, on my own. From programming to investing, I taught myself everything. I build tools that democratize access to financial freedom for consumers and businesses. My goal is to make it easier for people to learn how to make money, create software, and turn ideas into real products.
Have something to build?