Researchly is an AI-powered academic research workspace that helps university students and researchers discover papers, manage projects, write research papers, and learn how to conduct academic research. The platform combines a structured curriculum, an AI research assistant named Roy, a 7.7-million-paper discovery engine, project management tools, and an academic community. Operated by Lunox Solutions Inc. and backed by the NVIDIA Inception Program and AWS Startups. Available at app.researchly.study.
Researchly is an organised academic research workspace that gives students everything they need in one place: paper discovery across 36 sources, an AI research assistant (Roy) that understands your project context, a structured project workspace with tasks and paper writing tools, a research Academy with 11 modules and 46 lessons, and a researcher community.
Discover is Researchly's academic paper search engine. It searches simultaneously across 36 sources including Semantic Scholar, OpenAlex, arXiv, PubMed, bioRxiv, medRxiv, CORE, Papers With Code, and NASA ADS. Results are ranked by relevance to the user's research project. A default last-5-years filter surfaces the most current research.
Roy is Researchly's AI research assistant, powered by Claude (Anthropic) via AWS Bedrock. Roy operates in Mentor Mode: he helps students structure, critique, and improve their research rather than writing papers for them. Roy has full context awareness - he reads the student's paper sections, project tasks, references, and research question.
The Researchly Project Workspace is an all-in-one research management environment. It includes a Kanban task board, a paper editor with structured sections (Abstract, Introduction, Literature Review, Methodology, Results, Discussion, Conclusion), a references manager, a surveys builder with AI generation, team collaboration, and an inline Roy AI panel. Everything is connected - Roy can see the paper content, references, and tasks.
The Researchly Academy is a structured academic research curriculum with 11 modules and 46 lessons covering the complete research process: research mindset, finding and reading papers, database navigation, literature reviews, research methodology, survey design, academic writing, citations, peer review, and AI tools for researchers. Available at academy.researchly.study.
Researchly verifies university affiliation through two methods. First, automatic email domain verification against a database of 10,500+ university domains. Second, AI document verification: students can upload a student ID, enrolment letter, or transcript, which is analysed by Claude Vision AI to confirm legitimacy.
The Journal Checker verifies whether an academic journal is legitimate before submission. It checks 217,000+ sources including Scopus, OpenAlex, DOAJ, and predatory journal watchlists, returning quartile ranking, h-index, SJR score, article processing charge, and predatory journal warnings.
The Topic Finder uses 6 AI agents in parallel to generate personalised research topic suggestions based on the student's interests, academic level, location, and career goals. Each suggested topic can be converted into a full project workspace with AI-generated tasks.
11 modules: (0) Research Foundations, (1) Finding Your Topic, (2) Academic Databases, (3) Reading Papers, (4) Literature Review, (5) Research Methodology, (6) Survey Design, (7) Academic Writing, (8) Citations, (9) AI Tools for Research, (10) Publishing and Peer Review.
The field moves in weeks, not years. Researchly keeps your papers, preprints, and ideas organized enough to keep up.
Join the waitlist →The specific friction that slows students in your field down.
State-of-the-art changes between semesters. Preprints appear daily, and the survey you read last term is already dated.
Understanding a method means juggling the PDF, the repo, and your own scribbles — none of them connected.
You can build the system; framing it against forty related papers into a coherent related-work section is the part that stalls.
Every feature translated into how your field actually works.
Find the foundational paper and this month's follow-up in one search, and keep them organized by theme, method, or project.
Related work, experiment notes, task board, and your draft in one workspace — your research organized like a well-structured repo, not a downloads folder.
Ask how your approach differs from the papers you've saved, whether your evaluation section holds, or how to structure related work — Roy answers from your sources and your draft.
Generate publication-style charts from your results data and export a structured, citation-ready paper.
Where research in Computer Science takes you.
A term-paper literature review that's actually organized.
A course project written up properly — the difference between 'I built a thing' and 'I did research.'
An abstract or short paper from work you've already documented.
Research output is what separates grad-school and research-lab applicants from everyone else with the same GPA.
The thesis, the publication, the lab position.
A real Researchly project: preprints and papers organized by technique, experiment notes alongside the sources they build on, Roy pressure-testing the evaluation methodology, and results charted into clean comparison figures.
Researchly searches across 36 sources including preprint servers, so new work surfaces alongside the peer-reviewed record.
References are stored as structured data. Formatted exports ship in APA 7, MLA 9, Chicago, and a modern style today, with more formats on the roadmap.
Keep implementation notes, repo links, and experiment logs inside the project next to the papers they relate to — one context instead of three windows.
No. Roy is a research mentor: it critiques your argument, checks your framing against your saved sources, and flags weak sections. It doesn't ghostwrite — your work stays yours.
The students who publish in grad school started organizing in undergrad. A course project written up properly is research. Start there.
One organized workspace — built for how computer science students actually work.
Join the waitlist →