Submit Research for AI & Peer Review
Fill out your paper or article details below. Ingestion is 100% open-access, cookie-free, and published under CC BY-SA 4.0.
The Shutri Manifesto
As the cost of generating research approaches zero, the limiting factor becomes expert review. The Shutri Manifesto defines a practical reference architecture for conducting open, AI-assisted research review using freely available tools.
Empowering Domain Experts to Launch Independent Peer Review Services
Shutri is built for domain experts, independent researchers, and specialized communities who want to start peer review services in their own domain of expertise. We operate deepDive as an open reference ledger and authors are welcome to submit research, but the true goal is to democratize the peer review process itself—enabling anyone to operate an open, AI-assisted review node in their own discipline.
Minimizes recurring infrastructure costs so that participation is limited only by reviewer attention, not financial resources.
Research, discussions, revisions, and published ledgers remain publicly accessible without paywalls.
Independent capabilities (mdIngest, DDMA, mdBook) can be replaced or adapted to any discipline's workflow.
AI accelerates discovery & synthesis while human reviewers remain strictly responsible for every published conclusion.
Reviewer contributions generate visible public artifacts (publications, podcasts, diagrams) constituting a verifiable, open record of careful intellectual work.
The Three Pillars of Open Peer Review
An end-to-end open-source ecosystem engineered for research Ingestion, Publication, and multi-modal Dissemination.
Research Intake (`shutri`)
Submissions enter via open GitHub Issue Management. Research undergoes initial deepMind AI (`dm`) evaluation, generating an open synthesis report in the Mempool for peer review collaboration.
- ▪ Open GitHub Issue Substrate
- ▪ Transparent AI Perspective (`dm`)
- ▪ CC BY-SA 4.0 Open License
Research Publishing (`mdIngest` & `deepDive`)
Powered by the open-source Rust engine mdIngest (`coolchain`) to sanitize KaTeX math and citations into deepDive (deepdive.shutri.com)—a high-performance PWA ledger governed by the Law of 21.
- ▪ Lossless Python Payload (`extract.py`)
- ▪ Absolute KaTeX Math Hardening
- ▪ Progressive Web App (PWA) Precache
Research Dissemination (`ddma`)
Driven by DDMA (DeepDive Media Automator). Uses word-level Whisper transcription and parameter-tuned `acrossfade` stings to output master podcasts and 740x740 Square Videos (<20 MB for Nostr relay compliance).
- ▪ 740x740 Square Video Standard
- ▪ Nostr Protocol Compliance (< 20MB)
- ▪ Mosaic AI Motion Graphics
The Journey of Knowledge
How a research draft moves from an initial submission into an immutable ledger block.
1. Submitter Draft (Staging)
Author submits research URL & thesis via native form. Ingestion is 100% free.
2. Mempool (AI Ingestion & Review)
Gemini & `dm` evaluate core logic, generate an initial perspective report, and publish to the Mempool.
3. Template (Human-AI Production)
Author & reviewer build multi-modal assets: NotebookLM podcast audio, DDMA 740x740 videos, and Mosaic infographics.
4. Chain (Vector DB Consensus)
Once 21 verified papers pass public review, the block is minted to the permanent deepMind Vector DB chain.
What is in the Name? — Shutri (शु-त्रि)
Derived from Sanskrit Shubh / Shu, representing virtuous intent, pure knowledge, and sacred open dialogue free from insular paywalls.
Represents the SMS operational trinity: Pillar 1 Intake (`shutri`), Pillar 2 Publishing (`mdIngest`/`deepDive`), and Pillar 3 Promotion (`ddma`).
Synthesizes direct oral knowledge transmission (Śruti) with structured algorithmic logic (Sūtra): Intake hears the raw inquiry → Publishing threads the immutable ledger → Promotion broadcasts audio podcasts & 740x740 square videos back into public sound space.