Academic Research

Hand over a topic and get back the papers that actually matter on it — ranked, summarized, and linked. Every entry is a real, retrieved record: authors, venue, year, citation count, and an open-access link where one exists.

What it does

• Returns a ranked shortlist of papers, not a raw result dump
• Summarizes each in a sentence or two, plus a line on why it earned its place
• Surfaces the survey that saves you a week, when one exists
• Traces a paper you have: what cites it, what it builds on, what's related
• Honors your constraints — last three years, open access only, top 20

What to send

The topic, in your words. Optionally: a seed paper (DOI, arXiv ID, or title), a researcher's name, or any filters worth stating.

"RAG for legal documents, last two years, open access only."

What you get back

One markdown bibliography, best paper first:

**1. [Retrieval-Augmented Generation for LLMs: A Survey](https://doi.org/…)**
Gao, Xiong, Gao et al. · 2023 · arXiv · 687 citations · [open access](…)
Surveys the naive, advanced, and modular RAG paradigms, and the
retrieval–generation–augmentation triad the field is built on.
→ Best entry point: the taxonomy most later work is written against.

Good for

• Entering a field cold and needing the ten papers to read first
• Checking whether your idea has already been published
• Building a related-work section without missing the obvious citation

Limits

Sources are the OpenAlex catalog and the arXiv API. Summaries come from abstracts and metadata, never full text — so no numbers pulled from tables. English-language papers only. Coverage is strongest in computer science, physics, and math; broad but shallower for clinical medicine. Thin topics are called thin, not padded.