Skip to content

Repository files navigation

MARA: the Multi-Agent Review Assistant

The problem — peer review is breaking

  • An explosion of scientific publications – 3.3 million articles were published in 2022 alone and increases yearly​ [1]
  • Weeks-to-months delays: the median end-to-end time can be 160 days to acceptance [2]
  • Editor & reviewer overload: 80 % of review requests go to the same small cohort of scholars [3]

In short, human editors are drowning in administrative and analytic checks; authors wait; reviewers burn out; good science stalls.

MARA: the Multi-Agent Review Assistant

Our project turns the slow, overloaded peer-review pipeline into a “living editorial board” of cooperating reasoning agents that deliver the same checks a journal editor needs, in hours instead of weeks, while giving every stakeholder a clear, auditable report.

Alt text

We performed a small-scale benchmark by retrieving papers from openreview.net and evaluating our system on them. The results are presented below:

How to use?

A dockerized application for reviewing academic papers, consisting of a FastAPI backend, React frontend, and PostgreSQL database.

Getting Started

Prerequisites

  • Docker
  • Docker Compose

Environment Variables

The application uses environment variables for configuration. You can customize them by:

  1. Creating a .env file in the project root with the following variables:
# Database Configuration
POSTGRES_USER=postgres
POSTGRES_PASSWORD=postgres
POSTGRES_DB=postgres

# Backend Configuration
ALLOWED_ORIGINS=http://localhost:5173

# Frontend Configuration
VITE_API_URL=http://localhost:8000

Running the Application

  1. Build and start the containers:
docker-compose up -d
  1. Access the application:

  2. Stop the application:

docker-compose down

Development Mode

The application is configured for development with hot-reloading:

  • Frontend changes will automatically refresh the browser
  • Backend changes will automatically restart the API server
  • Volume mounts ensure local changes are reflected in the containers

Architecture

  • Frontend: React/TypeScript with Vite
  • Backend: FastAPI (Python)
  • Database: PostgreSQL

Containers

  • postgres: Database service
  • backend: FastAPI application
  • frontend: React application

Troubleshooting

  • If you encounter database connection issues, ensure PostgreSQL has fully started before the backend attempts to connect.
  • For permission issues with volumes, check Docker's volume permissions.

References

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

2 watching

Forks

Releases

Packages

Contributors

Languages