OpenAI 2026 hackathon

Tokengeoji (토큰거지)

A playful AI community where people post requests and turn public, remixable results into a shared feed of experiments.

Solo project by JungCollin SUNGWOO · 0 likes · 0 comments

Archive position — measured, not model output

0 likes on Devpost

2,264 of the 7,856 archived projects have more likes, and 5,592 share exactly 0 — so this project's #7,323 place in the like-ranked listing is a tie-break inside that group, not a ranking.

Projects (log scale)

1
10
100
1k
10k
05,592
11,758
2285
3–4132
5–975
10+14

Likes on Devpost. ▲ marks this project's group.

Show the figures
LikesProjectsShare of archive
05,59271.2%
11,75822.4%
22853.6%
3–41321.7%
5–9751.0%
10+140.2%
Devpost like counts for all 7,856 archived projects, captured when this archive was built.

Executive Summary

What the company appears to be

Tokengeoji (토큰거지) is a self-reported Korean-first AI community platform where users post structured AI requests and share public, remixable results. The author describes it as a space for people to ask everyday AI questions, receive community-generated outputs, and engage with others' work through reactions, comparisons, and remixing.

What changed

During Build Week (July 14–15), the author extended the product by reframing prompts in everyday language, adding an "ask-situation picker", improving personal benchmark flows, and enhancing UI/UX for profiles, authentication, moderation, and login. The project also included improvements to Docker deployment, localization for Korean users, and UX contract testing.

The single most important open question

Is there evidence of user engagement or community activity beyond the author’s own development efforts? The description lacks any data on actual usage, retention, or adoption — only claims about functionality and design decisions.

Back to contents

What The Product Actually Is

The description states that Tokengeoji is:

  • A Korean-first AI request-and-result community
  • Where users can post structured requests with context
  • Add text, links, images, or videos as public results
  • React to, accept, or remix results
  • Build profiles around their contributions
  • Use a same-prompt benchmark lane for comparing model outputs

It is described as a community feed of experiments, where AI-generated content becomes public and shareable.

Inference The product appears to be a lightweight social platform for sharing and discussing AI-generated content, with an emphasis on openness, remixability, and comparison. It uses Next.js and Firebase Authentication, and includes some testing infrastructure (Vitest, Playwright).

Back to contents

Positioning & Claim Evolution

The author states:

  • The inspiration was that AI requests usually disappear inside private chats, even when they could help others.
  • Tokengeoji aims to turn those requests into public community prompts.
  • It is a playful AI community where people post requests and turn results into a shared feed of experiments.

During Build Week, the author:

  • Reframed the product around everyday AI asks instead of technical jargon
  • Added an ask-situation picker
  • Built a personal same-prompt benchmark flow
  • Improved public profile, system-host labeling, help, login, and moderation-facing flows

Inference The positioning has evolved from a basic MVP to a more user-centric platform focused on accessibility, community engagement, and structured comparison of AI outputs. However, no evidence is provided that this evolution reflects actual user feedback or market demand.

Back to contents

Target Customer & ICP

The description states:

  • The product is Korean-first
  • It supports a community of people who post AI requests and results

It does not specify:

  • Who the core users are beyond "people"
  • Whether there’s a defined persona or segment
  • If it targets individuals, teams, or creators

Inference The target customer appears to be Korean-speaking individuals interested in AI tools, possibly including content creators, students, or general users seeking help with everyday tasks via AI. However, no explicit ICP is defined.

Back to contents

Business Model & Pricing Evidence

The description states:

  • The first request is frictionless
  • Repeat posting and editing use a lightweight ad-gate ritual
  • This supports the service without turning it into a marketplace

There is no mention of:

  • Revenue streams beyond ads
  • Pricing tiers or monetization strategy
  • Paid features or subscriptions
  • Any commercial model beyond the ad gate

Inference The business model seems to be ad-supported, with an optional frictionless entry point for first-time users and a lightweight ad-gate for repeat actions. No further details are provided.

Back to contents

Technical & Delivery Signals

The description states:

  • Built with Next.js App Router, TypeScript, Prisma with SQLite, Firebase Authentication, NextAuth sessions
  • Uses server-rendered community feed
  • Covered by Vitest (domain logic) and Playwright (core journeys)
  • Codex was used for inspecting product decisions, scoping changes, and verifying release paths
  • Docker path was hardened
  • Interface is localized for Korean users ("earth-and-hanji workbench")

Inference The technical stack suggests a modern, scalable web application built with React/Next.js. The use of testing tools and CI/CD practices indicates some level of engineering maturity. However, no evidence of production scale or performance data.

Back to contents

Traction & Maturity Signals

The description states:

  • The product existed before the submission period
  • Build Week work was focused on meaningful product extension
  • The dated commit history separates this Build Week work from the pre-existing MVP

There is no mention of:

  • User growth or retention metrics
  • Active user base
  • Customer acquisition or engagement data
  • Product usage statistics
  • Any form of traction beyond the author’s own development

Inference No evidence of traction, adoption, or user behavior is provided. The project appears to be in early-stage development, with no indication of real-world usage.

Back to contents

Competitive Context

The description does not mention:

  • Competitors
  • Market positioning relative to existing platforms
  • Similar products in the AI community space

Inference No competitive context is evident. It’s unclear whether Tokengeoji competes with other AI communities, social platforms, or tools for sharing AI outputs.

Back to contents

Key Risks & Red Flags

Key risks and red flags based on the description:

  • No traction or user data: The product exists only as a self-reported idea and author-developed prototype.
  • Unverified claims: All statements are self-reported; no third-party validation.
  • Unclear monetization strategy: Only an ad-gate is mentioned, with no clarity on scalability or profitability.
  • Limited scope: The platform seems to be focused on one region (Korea) and one type of interaction (AI requests/results).
  • No evidence of community engagement: No mention of user-generated content, activity, or feedback loops.

Back to contents

Diligence Questions To Ask The Founders

  1. What is the actual user base or community size?
  2. How many users are actively posting requests or results?
  3. What is the conversion rate from request to result?
  4. Are there any metrics on engagement (likes, reactions, remixes)?
  5. What are the key challenges in scaling beyond the current prototype?
  6. How do you plan to expand beyond the Korean market?
  7. Is there a clear path to monetization beyond ads?
  8. What is the long-term vision for the platform’s growth and community?

Back to contents

Investment/Partnership Verdict

Not evidenced

The description provides no evidence of:

  • Revenue or financials
  • Customer traction or adoption
  • Market validation
  • Product-market fit
  • Team experience or track record

This is a self-reported prototype, not a tested product with users or revenue. The author’s own account describes a functional MVP and some enhancements, but there is no indication of real-world usage or commercial viability.

Confidence level Low

Next steps

If this were a due-diligence context, further investigation would require access to actual user data, engagement metrics, and possibly a live demo or early-stage product review.

Back to contents

Source

Submitted to the OpenAI 2026 hackathon on Devpost. Project home on DevPost.

The analysis above was generated by a language model from the project's own one-line description. It is not independent research and contains no verified traction, revenue or customer data.