Extriver Partners
From market signals to software, businesses and capital
A student-run research institution. We study markets and creators, turn what we find into products, operate the companies that come out of it and allocate capital with the same discipline.
Discover Extriver
“We don't stop at quickly shipping software that fits the market. We keep measuring it until the data says what to build next.”
Extriver Partners brings students from leading universities together to study markets, companies and software, then turns what they find into real products, creator-led businesses and investments.
Extriver’s Creator-to-CEO hackathon. Choose a creator, understand their audience, and build software their followers will actually use and pay for. $37,600 total prize pool, with spots for 100 teams.
We started as a few students asking why good research so rarely becomes something people use. Today the same research loop drives a financial terminal, a consumer portal, games, creator products and a verified investment track record.
Find out more about usResearchWe study markets before they move.Market structure and sizing, consumer behavior and willingness to pay, and the audiences behind creators: who follows them, where they live, what they buy and why. Quantitative research, from factor models and ML-assisted valuation to sentiment and capital-flow analysis, turns each question into numbers we can test, publish and defend.
BuildWe turn opportunity into software.Every gap the research uncovers becomes something we can ship: consumer apps, financial tools and creator products built around real demand. Analytics and CRM are wired in from day one, so every signup, purchase and churn signal flows back into the research loop and tells us what to build next.
InvestWe put capital behind conviction.When the evidence is strong enough, we invest directly in the market and consumer theses we have studied, back the ventures that grow out of the opportunities we find, and research listed equities with the same discipline. Every position is documented, measured and reconciled, so conviction is earned long before it is expressed in capital.
Learn by doing real work.
Research that gets published, code that ships to real users, and launches with creators, alongside members from 13 universities.
Scale your audience with us.
We analyze your followers, design software with you, and keep measuring after launch, so your audience becomes a business you own.
Join our mission of turning research into real value.
Are you a company, fund or foundation that wants to support student research and the products it creates? We would love to hear from you and discuss how we can work together.
Learn more50 members
+ more
Are you a student who wants to do real research, ship real software and learn from members across 13 universities? Join a team that turns markets into products.
Find your role
Is there a market, company or question you want studied? Tell us what you need and our research team will reply with how we would approach it.
Tell us your question
Are you a company, fund or foundation that wants to support student research and the products it creates? Reach out to us for a discussion.
Partner with us
Are you a creator who wants your audience to become a business? We analyze your followers and design software together with you.
Our productsWe don't stop at quickly shipping software that fits the market.
After launch, we keep monitoring each product's data and financials to support its operations and growth.
The same quant and market-regime research also drives our Venture Capital and Public Equity Investment.
We combine Quantitative Research, Software Engineering, and Business Management into one platform that creates products, operates companies, and allocates capital.
Real-time news and industry data, combined with trend analysis and quant models, tell us the current market regime.
From consumer behavior, industry growth, competition, capital flows, and structural shifts, we find products with real demand in today's market, not just trends.
Each opportunity becomes new software, a data product, or a digital business.
We match each product with creators whose followers fit its customers.
Creators don't just promote. They run branding, marketing, customer communication, and operations as their own small business.
We provide the technology and operations that turn their Audience, Trust, and Distribution into real, lasting revenue.
We guarantee continuous monitoring after hand-off. Revenue, conversion, acquisition cost, retention, marketing efficiency, costs, and cash flow are analyzed automatically to forecast performance.
When something goes wrong, we find the cause in the data and propose the fix.
Researchers from SKY and Ivy League universities support creators with industry, competitor, and financial analysis and strategy.
We quickly build what research finds: web services, AI tools, data platforms, and automation, each shaped to the customer's problem.
Usage and operating data keep refining features, pricing, user experience, and the business model.
The result: products built to grow over the long term, not just MVPs.
Creators' existing communities are our first distribution network.
Follower, content, purchase, and conversion data show which creator and product pairs will perform best.
Instead of big ad budgets, we run our own growth engine: more growth brings more data, and better next launches.
By analyzing industry growth, earnings, valuation, regimes, macro variables, and capital flows, our systems find investment opportunities in both private and public markets.
Early-stage opportunities lead to venture capital investment; analysis of listed companies leads to equity investment.
New industries reveal companies, companies reveal new markets, and in those markets we build again.
We analyze markets and companies with data.
We turn the opportunities we find into real products and systems.
We distribute and operate products, and allocate capital again.
We are building one integrated system that analyzes markets, builds products, operates companies, and allocates capital.
A software and business research society.
At its core, Extriver Partners is a research society. Our members come together to study markets, companies and software, and to share that work through published research.
Research published by Extriver Partners across four disciplines.
Four research areas, one question behind each: where is real value forming before the market prices it in?
We focus on models that still work on data they have never seen: regime detection, factor exposures, and machine-learning forecasts that are tested across thousands of companies before any of them is trusted.
Tell us the market, company, or question you want studied. Our team will reply with how we would approach it.
Research in this category will be published here.
Research becomes software. We build the products, the infrastructure they run on, and the distribution that launches them.
Products we build and the projects we develop for partners.
FintechIn-house productAll-in-one financial terminal for Korean investors: quotes, news, company research, portfolio, backtesting, quant, derivatives, and macro indicators in Korean, in one place.
arterm.xyz ↗
Consumer platformIn-house productKorean product and culture portal: free ingredient analysis and verified review data for Korean products, monetized through brand listings and banner ads.
Represented by LoveSoo ↗
AI · ConsumerIn-house productAI assistant for parents: an age-adapted 28-question child temperament assessment, 40 analysis cards, and LLM guidance grounded in PubMed references, in four languages.
GamesIn-house productGames built in-house, including ECLIPSE, a 50-level browser pixel action RPG with real-time 1v1 PvP, and the Extriver Story Minecraft RPG server.
Represented by @superdupercatai ↗
GamesFull developmentHacketmon, a Minecraft streamer server on SOOP (Sep 1–4, 2026) played by Gamst. Developed end to end by Extriver.
Watch Gamst on YouTube ↗
GamesDevelopment contributorDOS Online (dose.kr), a long-running Korean Minecraft real-economy server. Extriver took part in development.
Official DOS Online café ↗We are partnered with more than ten influencers. Their audiences are our first distribution channel, measured and matched to each product before launch.
Freshly squeezed AI comics every day! Launched by Extriver.
Instagram bio
I'm Soo, a 50-year-old Korean woman living in the United States, married to a younger man, and starting a whole new chapter of my life. Here, I share my everyday life: Korean cooking, cozy meals, fitness, marriage, friendships, and honest conversations about healing, growing older, and finding strength again. Life doesn't stop at 50.
YouTube channel description
Dream. Believe. Achieve. Artist, choreographer and performer.
YouTube channel description
Indian channel on online side income, money apps, rewards and digital finance.
Follower and post counts as of October 2026.
Three product lines built on one research and software stack.
Factor, regime, market, and valuation models delivered as usable tools.
The same models used for our own research and investment decisions, packaged with clean data access and documented methodology.
Market and audience research translated into tested business systems.
Tools that size markets, compare business models, and project unit economics before a product is built.
Products launched, distributed, and improved through creator and influencer networks.
Software matched to creators whose audiences fit the customer profile, with monitoring that continues after launch.
How research runs as production software.
Everything starts with data: market prices, company filings, creator audiences and product usage. Pipelines collect it automatically, clean it and keep it current, so every model works from the same reliable source.
01 Data → 02 Pipeline
Research turns that data into models: forecasts, factor and regime signals, valuations. Each model is served through an API, so every product uses the same tested logic instead of rebuilding it.
03 Model → 04 API
Our terminals, apps and creator products call those APIs and put the results in front of real people, from investors to parents to gamers.
05 Software → 06 User
How people actually use the software flows back as data. We measure what works and what does not, and feed it into the next round of models and features.
07 Feedback → 08 Optimization
Research runs on production software: data pipelines, models, APIs, and applications connected in one feedback loop.
Each stage activates the next. Every cycle starts with more data than the last.
Find a measurable opportunity.
Quantify demand, pricing, and economics.
Test the smallest working version.
Build the production system.
Launch through creator networks.
Collect usage and financial data.
Feed results into the next cycle.
Examples of what the system produces.
Monitor regimes, factor exposures, and signals in one view.
Compare business models for a market before building.
Match products to creators and track performance after launch.
Module names and specifications are illustrative.
Not every market we find becomes software.
Some opportunities are best captured as capital. When our research finds a company before the market does, we back it as a venture investment; when it finds mispriced value in the public markets, we take a position in listed equity.
Venture investments and listed equity positions selected through the same research system that drives our products.
Early-stage companies identified through market, product, and business-model research.
An early-stage company identified through the market and product research process that also drives our own software. The investment reflects the crossover between our product research and venture capital.
Sector, business model, and thesis details for Afformation to be added.
Listed companies selected through quantitative, regime, macro, and financial analysis.
Total value created from ₩65.9M of starting capital: ₩176.0M of profit, of which ₩85.0M has already been withdrawn.
The position was built around the memory cycle. Inventory, contract pricing, and AI-related demand indicators pointed to a recovery before it was reflected in consensus earnings.
Accumulated 1,000 shares at ₩55,400–57,200 between November 2024 and January 2025, then scaled out as the stock re-rated to ₩354,000. 400 shares are still held at an average cost of ₩182,496.
Shipbuilding upcycle driven by fleet replacement and LNG carrier demand, identified through regime and order-backlog analysis.
Memory-cycle recovery positioned with inventory, pricing, and AI-driven demand indicators.
Strategic-metals exposure and ownership-structure events evaluated with financial and scenario analysis.
Mid-size tanker builder within the same shipbuilding regime, sized as a secondary cyclical position.
All figures reconcile to Kiwoom Securities transaction statements (Oct 2021 – Oct 2026; prices as of 12 Oct 2026). Return is profit divided by the cost of shares sold, net of commissions and transaction tax; profit includes dividends and, for open positions, unrealized gains. Also traded: Nota (486990 · KOSDAQ), an IPO round-trip in Nov 2025 for +₩0.1M. USD figures converted at ₩1,400.
Across New York, Boston, Waterloo, Seoul, and Hong Kong.
Institution direction, capital allocation, and governance.
Founded Extriver Partners to connect quantitative research, software engineering, and business operation in one system. Leads the institution's research direction and engineering.
Research and engineering leadership across the institution.
Models, data infrastructure, and the software systems that run our research and products.
Quantitative, market, industry, and economic research across asset classes and sectors.
Creator and business partnerships across products and portfolio companies.
Outside researchers who work with Extriver on specific projects.
The institutions, sponsors, and teams we work alongside.
From AI workshops with MongoDB to the sponsors and studios who build with us, every partnership connects our research to the real world.
Organizations we collaborate with on research.
Held an educational workshop with MongoDB on Vector Search and its applications in AI-powered applications.
Supporting builders at Hack the Audience 2026.
A domain name ending for the next generation of internet users, with flexible and affordable names for every website. Every Hack the Audience participant gets a .xyz domain free for 1 year.
An open-source marketing system your coding agent can run. You build the product, and Tin's AI helps you win customers and run your marketing.
A cloud application platform. Push your code to GitHub and Render deploys and runs your web services, databases, and AI agents 24/7.
Teams we have built projects with.
DOS Online (dose.kr), a long-running Korean Minecraft real-economy server. Extriver took part in development.
Hacketmon, a Minecraft streamer server on SOOP (Sep 1–4, 2026) played by Gamst. Developed end to end by Extriver.
Sponsored Instagram Reel collaboration with Tripo AI, the AI 3D model generation platform, arranged with the global marketing agency Tec-Do for SuperDuperCatAI.
Discover what you can build, and find your team.
We hire across research, engineering, and business. Tell us what you would build.
We look for researchers, engineers, and operators who can think rigorously and build systems.
See open rolesFrom your first research project to your first shipped product, there is room to grow while you study.
University students and recent graduates from around the world, in research, engineering, design or business.
Hands-on experience across research, software and business, a public record of your work, and members who have been where you are.
Already shipping code, models or businesses? Lead the work from first idea to post-launch data.
Engineers, researchers and operators who have already built and shipped real work.
Ownership of a whole product or research line, and a team that connects quantitative research, software and business in one loop.
A student-run research society: we study markets, build software and share what we learn.
Anyone who wants to think rigorously, build systems and work with people across universities and countries.
A place to learn by doing real work, with colleagues from many universities, and a portfolio you can show.
Grouped by function. All roles work across research, software, and business.
Develop systematic models, regime analysis, and market research tools.
APPLY →Research industries and markets to identify products and investments.
APPLY →Build forecasting and classification models that run in production.
APPLY →Build data pipelines, APIs, and the applications our products run on.
APPLY →Match products with creators and support them as business operators.
APPLY →Monitor launched products and turn operating data into improvements.
APPLY →Five qualities we evaluate in every process.
01Claims are tested against data before they are acted on.
02You follow questions across markets, technology, and business.
03You can read a model, write the code, and explain both.
04You carry work from research to production to results.
05You see how research, software, and capital connect.
Four steps. We respond to every application.
Submit your background and what you would build.
We review every application against the role.
A research problem or technical exercise relevant to the role.
A conversation with the founder and the team you would join.
