A philosophy graduate wanted a one-click homepage for his papers. Eighteen years later, Academia.edu has turned that small irritation into a vast research network - and a business built around the desire to know who is reading.
The London startup wanted to map the world’s scientific reading. Nineteen years, one controversial sale and several retired products later, Mendeley has learned that researchers mostly want their PDFs tidy and their citations correct.

Before AnswerThis became a research platform used around the world, it was a teenager’s refusal to keep downloading papers one by one. Ayush Garg has spent years turning that small irritation into a larger argument: scientific work should move at the speed of the questions, not the software.
U of T began as an Anglican college that took 16 years to open and lasted only seven years in the classroom. Its secular reboot became a three-campus platform serving 103,140 students - and a case study in how institutions turn breadth into an advantage without pretending complexity is free.
The simple, quick and anonymous online survey tool that's bursting with juicy insights.
Most survey companies can read your answers. BlockSurvey built one that can't - end-to-end encrypted forms with an AI agent on top, run by two ex-fintech engineers who won a hackathon and never went back.
AnswerThis is an AI research assistant that turns literature reviews from months of work into minutes. It searches a database of more than 250 million academic papers and answers complex scientific questions with line-by-line, verifiable citations. Founded by University of Richmond seniors Ayush Garg and Ryan McCarroll and backed by Y Combinator (Fall 2025), the platform is used by over 250,000 researchers and crossed $1M in annual recurring revenue within eight months.
Latent Knowledge is a New York-based AI company building LitView, a semantic search and literature-visualization tool that helps R&D teams and academic researchers discover, cluster, and map scientific literature beyond simple keyword matching. Founded in 2019 by James Reilly, the company uses applied natural language processing to surface hidden links between technical articles and speed up the research process across public and private data sources.