From late-night ramen to launch day: Rutgers freshman Alan Cheng built PubSort to crush academic jargon into bite-size insights. Now the 18-year-old invites the world to test the AI that can slash weeks off a lit-review.
NEW BRUNSWICK, N.J., April 24, 2025 /PRNewswire-PRWeb/ -- PubSort, a free website created by 18-year-old Rutgers freshman Alan Cheng, is now live with a mission to make scholarly research instantly understandable. The platform compresses the key findings of more than 500,000 peer-reviewed studies into plain-language summaries that can be read in under 15 seconds, helping students, journalists and professionals decide at a glance whether a paper merits deeper reading.
After struggling through jargon-laden PDFs while in his first year at Rutgers as an engineering student, Cheng began coding PubSort in his dorm room. By mapping every paper into dense vector embeddings, PubSort scans a half-million studies in milliseconds, turning hours of keyword scrolling into a single, context-aware query.
How PubSort Works
- Semantic search accepts natural-language questions and surfaces the most relevant studies, even without exact keywords.
- Each paper receives a one-sentence summary and short segments covering hypothesis, methods, sample size, results, and limitations.
- A data-quality meter warns users when a study is retracted or statistically under-powered, drawing on the Retraction Watch database.
- Summaries and citation-ready details can be copied instantly, making it easy to drop key findings into literature reviews, dashboards, or class projects.
Early Testing Results
In controlled trials with 300 undergraduates across Rutgers, Cornell, and the University of Pennsylvania, PubSort shortened literature-review time by 68 percent and improved citation accuracy by 41 percent. The tool has since been adopted by study-group leaders and university peers to streamline background reading for lab assignments.
Roadmap for 2025
- Q3 2025: Browser extension that overlays PubSort summaries directly onto publisher websites.
- Q4 2025: Multilingual abstracts in Mandarin, Spanish, and Arabic to broaden global accessibility.
- Ongoing: Partnerships with open-access journals to embed an "Instant Summary" button alongside newly published articles.
Get Involved
PubSort is actively seeking feedback from early users and welcomes collaborations with educators, open-science advocates, and developers. Feature requests and bug reports can be submitted through the site's feedback portal, while API keys for academic projects are available upon request.
About PubSort
PubSort is a free, AI-powered academic discovery platform founded in 2025 by Rutgers student Alan Cheng. By combining semantic search with large-language-model summarization, PubSort turns dense research papers into quick, trustworthy insights that anyone can understand and reuse. Visit www.pubsort.com for more information.
Media Contact
Alan Cheng, Pubsort, 1 8185708801, [email protected]
SOURCE Pubsort
Share this article