AcademiaX Press is building a modern, digital-first press where publishing is treated as a repeatable, auditable workflow rather than a bespoke, opaque process. Our goal is simple: publish high‑quality scholarly content faster—without sacrificing rigor—by making the workflow explicit, policy-driven, and measurable.
What we publish
We publish research articles, news, opinion, analysis, podcasts, videos, and books. Not every content type is peer reviewed: research articles follow the strictest pipeline, while news and commentary are editorially reviewed for clarity and accuracy.
- Research articles: scholarly articles published with structured provenance, revision logs, and print-friendly HTML/PDF artifacts.
- News & analysis: platform updates, policy discussions, and deep dives into reproducible, AI‑assisted publishing workflows.
- Books: long-form, versioned releases (digital-first) with the same evidence discipline as our research pipeline.
Journals vs books (digital-first)
Journals are organized into volumes and issues. Books are organized into editions and series. For digital-only releases we do not require standard identifiers, but when a print edition exists we support ISSN/ISBN metadata for external indexing.
Explore our journals and our books.
Our editorial model (agents + policy)
Our “editorial board” is a set of specialist agents that run discovery, drafting, critique, validation, and production. Every step emits artifacts (drafts, tables, logs) and is constrained by explicit rules.
- Discovery identifies topics and sources using transparent queries and inclusion criteria.
- Drafting is citations-first: the system is not allowed to introduce uncited claims.
- Peer review produces structured critiques and bounded revision loops (no infinite cycles).
- Validation attaches evidence traces to claims and flags unsupported statements.
- Production renders semantic HTML and camera‑ready PDFs with stable artifact links.
Our “editorial board” is a set of specialist agents that run discovery, drafting, critique, validation, and typesetting. Learn more about our agents.
Provenance and reproducibility
“AI-assisted” must not mean “unverifiable”. We treat provenance as a first-class product feature: citations, evidence traces, and versioned artifacts are what make results inspectable by readers, reviewers, and institutions.
- Evidence discipline: claims are grounded to verified sources; weakly-supported claims are removed or flagged.
- Versioned artifacts: drafts, tables, and PDFs are stored with immutable identifiers.
- Transparent review history: critique logs and decision checkpoints are preserved.
Policies
We enforce citations-first generation and provenance capture across the pipeline for every journal and book. Read Editorial policies.
Open access, funding, and institutions
We aim to keep reading friction low and publishing transparent. Costs are driven by compute and verification workloads—not by arbitrary tiers. Institutions can request plans for sponsorship or subscription support.
Where to start
- Want to publish? Start at Publish with us.
- Want to understand the workflow? Read our agents and research taxonomy.
- Want practical guidance? Use the Guide for authors & referees.
- Want transparency on operations? See Platform metrics and /api/healthz.
Frequently asked questions
Is everything written by AI?
No. We publish multiple content types. Research articles and technical posts may be produced with agent assistance, but we distinguish between peer‑reviewed research and editorial content. All content is expected to meet provenance and policy requirements appropriate to its type.
Can humans participate in review?
Yes. In enterprise configurations, human referees can co‑review alongside Critic agents. Reviews can be stored as structured logs for transparency.
How do you prevent hallucinations?
We enforce citations-first drafting, DOI verification where applicable, and validation checks that attach evidence traces to claims. Unsupported claims are removed or explicitly flagged for author attention.