You make the decisions.
Choose the gap, proofread the paper, and approve each submission.
A clear path through your research.
Choose a gap
Compare evidence-backed questions. Choose one to pursue.
About gap discovery
Start with a topic, domain, or idea. The agent compares potential contributions with the closest existing work, available evidence, and limitations. You select the question to pursue.
Explore a research ideaFollow the evidence
Inspect the sources behind each finding.
About the research
Searches, screening decisions, and extracted passages stay with your project. Source-access limits and contradictory findings remain visible. You can optionally review the plan and sources before drafting.
Explore the sourcesReview your draft
Check the claims and refine your manuscript.
About drafting and review
Develop a review, conceptual framework, or reasoned argument from published evidence. Inspect citations, accept or reject revisions, and correct blocking issues. You proofread the manuscript before approving it.
Find your paper typeFind a journal
Compare publishing routes. Approve each submission.
About publishing
Explore suitable, verified no-fee external journals or the planned discipline journals from Autonomous Research Publishing. You decide where to pursue publication. Journal acceptance is not guaranteed.
Compare publishing routes
How we’re different.
Methods & qualityInspect the evidence.
Examine sources, supporting passages, contradictory findings, and access limitations.
Prepare for publication.
Develop the manuscript and compare publishing routes, with acceptance decided by the journal.
Explore the sources
Databases, open-access papers, and your own documents.
Crossref
Find published papers and their DOIs.
OpenAlex
Explore related research and open-access locations.
PubMed
Search biomedical literature when relevant.
Open-access sources
Read available full texts, with access limits recorded.
Your source library
Add PDF, text, and Markdown sources.
Official journal policies
Check scope, fees, peer review, and AI policies.
PAPER TYPES
Supported paper types.
Six ways to build on existing knowledge.
Narrative review
Best for: Understanding a topic and connecting themes across studies.
Critical review
Best for: Evaluating the strengths, weaknesses and disagreements in existing work.
Conceptual paper
Best for: Clarifying ideas or building a framework from existing theories.
Argumentative paper
Best for: Defending a position on a debatable question.
Systematic review
Best for: Answering a focused question through a reproducible literature review.
Scoping review
Best for: Mapping what has been studied and where evidence is missing.
EXPLORE AN EXAMPLE
See how a question becomes a paper.
Illustrative questions and review steps. Each proposed contribution still needs an evidence assessment.
Conceptual paper
When does transparency support accountability in AI systems?
Possible contribution: Distinguish disclosure from answerability, and propose conditions connecting them.
Evidence needed: Compare foundational arguments about transparency, institutional responsibilities, and auditing.
A proposed conceptual comparison; novelty and practical effectiveness are not established.
Customer stories
“I had a topic, but not a researchable question. Comparing the proposed gaps helped me see what a literature-based paper could actually contribute.”
Illustrative researcherGraduate student in education
Social sciences
From a broad interest in feedback to a focused review question
Read this example
In this fictional scenario, a student explores published accounts of feedback and learner agency. The agent proposes bounded questions and explains their evidence needs. The student chooses a question, reviews the plan, and remains responsible for checking the eventual manuscript. This example does not establish that the question is novel.
“The useful part was seeing where the sources disagreed. I could inspect the passages, challenge the interpretation, and decide which claims belonged in my argument.”
Illustrative researcherIndependent author in technology ethics
Humanities
Building an argument that makes room for disagreement
Read this example
This fictional author develops an argumentative paper from published AI-ethics literature. Competing interpretations remain visible beside the draft. The author checks source passages, revises the premises, and decides whether the rebuttals are warranted. No empirical study or publication outcome is represented.
“I wanted a draft I could question, not just read. Having the evidence beside the manuscript made it easier to check the scope of a claim before accepting a revision.”
Illustrative researcherResearcher in environmental science
Natural sciences
Keeping the evidence in view while refining a literature review
Read this example
In this fictional narrative-review workflow, an author compares published findings about urban vegetation and air quality. Access limits and contradictory results remain part of the record. The author proofreads the draft and evaluates each proposed revision. The example does not imply validated findings or journal acceptance.
ACADEMIC INSTITUTIONS
Academic institutions.
- Harvard UniversityACADEMIA
- Duke UniversityACADEMIA
- Northwestern UniversityACADEMIA
- Boston CollegeACADEMIA
- Columbia UniversityACADEMIA
- Northeastern UniversityACADEMIA
- University of North Carolina at Chapel HillACADEMIA
- University of MiamiACADEMIA
- Loyola University ChicagoACADEMIA
9 of 9 institutions
Good questions.
What kind of research can I do?
Work with existing knowledge: narrative and critical reviews, conceptual and argumentative papers, and appropriately scoped systematic or scoping reviews. The workflow does not run experiments, recruit participants, collect original empirical data, or perform meta-analysis.
How can I check the evidence?
Your project keeps its sources, search history, screening decisions, and extracted passages together. Open a source to inspect the text behind a claim. Abstract-only access, incomplete coverage, and unresolved questions remain visible for your review.
What do I need to do?
Choose the research gap you want to pursue, then carefully proofread the manuscript and check its sources. You provide author details and declarations, and explicitly approve the final submission package. You remain responsible for the paper.
How are journals selected?
The large-publisher route checks official policies for scope, peer review, fees, article types, and AI use. Only verified no-fee journals qualify; documented acceptance rates are historical, and acceptance is never guaranteed. Autonomous Research Publishing is a planned in-house alternative, with discipline-specific journals and an initial human-feedback target of one business day. Its editorial service and submissions are not open yet.
Can I bring my own sources?
Yes. Add PDF, text, or Markdown sources to a project. The agent can combine accessible source text with the scholarly literature it retrieves. Source-access limitations stay attached to the evidence.
Is Autonomous Research free to use?
The beta currently includes 2 new projects per rolling 30 days, with limits on research tasks. Your projects are private, and you can export your manuscripts and research records.