Openhouse wants to become the largest researcher community in the world. It also aims to be the first community in human history with over ten thousand members that is fully managed by AI.
Our idea originated from a highly - upvoted suggestion under a post made by hundreds of neuroscientists. They wanted to conduct in - depth research on the latest neuroscience studies combined with algorithmic emotions to drive the development of human technology. We are determined to carry on this initiative and plan to reach out to tens of thousands of scholars within three months.
THE SAGE is an AI agent designed to manage the overall operation of the community. Currently, we have developed two key functions:
Contribution Evaluation: SAGE automatically assesses each user's contribution to the community based on the content they post, their areas of expertise, and their industry. Similar to GitHub, it records each person's daily contributions.
Reward Distribution : SAGE automatically distributes rewards based on individual contributions. For example, users who create highly - liked or widely - bookmarked high - quality content will receive rewards. Individuals will be able to import their past contributions, such as GitHub projects and published papers to earn a substantial amount of virtual currency with just one click.
OH Mate is different from traditional social matching. It is a newly designed Slow - Matching Mode. This mode is used to match potential scientific researchers or unique combinations may have amazing synergies. For instance, it could match a neuroscientist with an algorithm engineer, or a philosopher with an oceanographer. The matching process may take several days. OH Mate will continuously search for truly remarkable connections for you. The more information you provide, the better the potential matching results are likely to be.
OH believes that both research and learning are forms of contribution. It aims to provide a platform for users around the world who are willing to participate in human development or learning without discrimination. This includes 235 million researchers above the university level globally (data from UNESCO), and 1.81 billion high school/middle school students. Through AI Sage, people at any stage can easily get involved in projects suitable for them and be matched with OH mates.
- Engineer Smith and biologist Annie met on the platform, and together they developed a machine capable of cave rescue by studying the movement of squids.
- High school programming enthusiast Constance discovered Andy's new algorithm article on the platform and, through hard work, joined Andy's research team.
- Professor Andrew found a user named Liu on the platform. After reviewing Liu's personal Sage profile (showcasing his research), he saw that Liu had been diligently working on eco-friendly technologies for two years with a very innovative approach, so he invited Liu to collaborate on an experiment.
Johnny Lapaca Peking University Product Designer
JianJian Tsinghua University Frontend/ai Engineer
Charlie Columbia University Backend/ai Engineer
Eva Cornell University UI
π Demo: https://openhouse.horik.cn
| Layer | Technology |
|---|---|
| Frontend | React + Vite + TailwindCSS |
| Backend | Go + Gin + GORM + MySQL |
| Database | MySQL 8.x |
| Auth | Email, GitHub, Google OAuth2 |
| AI Match | LLM API (OpenAI / TogetherAI) |
| Storage | Alibaba Cloud OSS (Image CDN) |
- Passwordless login via email verification code
- OAuth2 login via GitHub & Google
- JWT-based authentication & authorization
- Support for multiple account bindings (e.g., Email + GitHub)
- Editable user profile (nickname, gender, avatar, intro)
- Avatar uploaded to OSS
- Track bound login methods (email/github/google)
- Create, edit, and delete posts with text and images
- Like, favorite, comment on posts
- Public feed with follow-based filtering
- Anonymous βTree Holeβ mode (optional)
- AI-based scoring to manage the community (future implementation)
- Users submit tags, intro, and research area to join match pool
- Daily LLM-powered intelligent matching
- Matches scored with AI comments and reasons
- Results revealed once per day
- Matching statuses:
Not Applied,Matching,Matched,Revealed
- One-on-one chat unlocked after successful match
- Polling-based new message retrieval
- Structured schema with sender/receiver UUID tracking
- System notifications: match success, likes, comments, admin messages
- User messages: one-on-one chat after match
.
βββ backend/ # Go + Gin backend
β βββ api/ # API layer
β βββ model/ # request/response/database models
β βββ service/ # business logic
β βββ middleware/ # JWT auth, logging
β βββ utils/ # helper functions
β βββ global/ # global variables
β βββ initialize/ # DB, OSS, config init
β βββ main.go # project entrypoint
βββ frontend/ # React + Vite frontend
- Go 1.20+
- MySQL 8.0+
- Node.js 18+ (for frontend)
- Docker (optional for local deployment)
Backend:
cd backend/
go mod tidyFrontend:
cd frontend/
pnpm installcd backend/
go run main.goβ Swagger Docs: https://openhouse.horik.cn/swagger/index.html#/
cd frontend/
pnpm devThis project is licensed under the Apache-2.0 License.
See the LICENSE file for details.
We welcome contributions from researchers, developers, and designers.
To contribute:
- Fork this repository
- Open a pull request
- Or reach out via GitHub Issues
Let us build the worldβs largest researcher community β together.