Paperzilla turns scattered paper sources and forwarded paper alerts into a relevant, always up-to-date research feed, delivered to your inbox, app, RSS reader, CLI, MCP client, or AI agent.
* Papers and associated software repos and datasets.
System overview
From research-data chaos to focused research context
Paperzilla brings together papers, datasets, software repositories, patents, and forwarded alerts, filters them against the research topics you care about, and turns them into structured feeds for researchers, teams, tools, and AI agents.
Inputs
Relevance Engine
Outputs
Inputs
Papers, alerts, datasets, software, patents
arXiv
preprint
bioRxiv
preprint
medRxiv
preprint
ChemRxiv
preprint
Scholar Alerts
email
Journal Alerts
email
Custom Sources
custom
Datasets*
data
Software Repos*
software
Patents*
patents
Paperzilla
Relevance Engine
Topic filter
Scoped to your research interest
Relevance ranking
Strongest matches rise first
Feedback loop
Learns what matters over time
improves over time
Outputs
Feeds and access for people, tools, and agents
Feed
App Feed
human
Mail
Email Digest
human
RSS
RSS / Atom
human
MCP
MCP Server
agent
>
CLI Access
agent
API
API Access*
agent
1
Connect sources
2
Tune relevance
3
Get focused feeds
Scattered papers, datasets, repositories, patents, and alerts go in. Paperzilla filters for relevance. Structured research feeds come out, ready for humans and AI agents. Items marked with * are paper-related datasets, software repos, patents, and API access.
How it works
How Paperzilla works
Paperzilla continuously monitors research sources, filters incoming papers and research data against your topic, and delivers the results where your research workflow already happens.
01
Input
Connect sources and alerts
Harvest papers from preprint servers, bring in paper-related datasets, software repos, and patents, and forward existing alerts into Paperzilla. Bring the scattered streams into one place without rebuilding your monitoring setup from scratch.
02
Relevance
Filter and rank by relevance
Define the topic you care about. Paperzilla ranks incoming research objects, separates must-reads from background noise, and improves over time with feedback and recommender-style signals.
03
Output
Send structured output anywhere
Read your feed in the app, receive daily or weekly email digests, subscribe with RSS, or connect agent workflows through MCP, CLI, and API access.
Sources and alerts
Bring every research stream into one relevance filter
Paperzilla can monitor preprint sources directly, accept forwarded alerts, and fold paper-related datasets, software repos, and patents into the same feed.
Paper sources
Track new papers from major preprint servers without manually checking each source.
arXiv
bioRxiv
medRxiv
ChemRxiv
ChinaXiv
Forwarded alerts
Forward existing alerts to Paperzilla and let the relevance engine decide what belongs in your focused feed.
Google Scholar alerts
Paper alert emails
Journal alerts
Newsletter-style alerts
Research data sources
Paper-related datasets, software repos, and patents fit into the same relevance-filtered workflow.
Datasets *
Software repositories *
Patent records *
Additional sources on request
Items marked with * are paper-related datasets, software repos, and patents. Need another source? Let us know.
Built for humans and agents
Built for researchers, teams, and AI agents
Paperzilla does not lock your research data feed into one interface. The same relevance-filtered stream can be read by people, shared with teams, or consumed by AI agents and automation workflows.
Human review
For researchers and teams
Skim the app feed, receive daily or weekly email digests, or subscribe through RSS. Stay current on narrow research topics without living in alerts, search pages, and source-specific inboxes.
In-app research feed
Daily or weekly email digest
RSS/Atom feed
Topic-specific monitoring
Must-read and related-source triage
Agent workflows
For AI agents and automation
Give research agents a current, scoped research-data stream instead of asking them to rediscover the field from scratch. MCP, CLI, and API access make Paperzilla output usable inside agent workflows, scripts, and recurring research briefs.
MCP server access
CLI access
Agent-ready research streams
Structured research context
Recurring briefing workflows
Use cases
A focused feed for every research workflow
Use Paperzilla anywhere missing the right paper, dataset, repository, or patent would slow down a project, weaken a decision, or leave an agent working from stale context.
01
Researchers
Track a narrow topic without checking every source, alert, and newsletter manually. See the papers, datasets, repositories, and patents most likely to matter before they disappear into the pile.
02
Labs and research teams
Keep projects, proposals, journal clubs, and student work aligned with the newest relevant papers, datasets, repositories, and patents.
03
R&D teams and founders
Monitor technical areas that could affect roadmap decisions, product direction, competitive analysis, scientific claims, or patent exposure.
04
AI research agents
Feed agents a continuously updated, topic-specific research-data stream through MCP, CLI, or API workflows, so they start from focused context instead of the open web.
Example feeds
Example research feeds
Explore example feeds to see how Paperzilla tracks different research contexts, then use one as a starting point for your own topic. Browse the public digests for more.
Computer vision, machine learning, AI
Vision and multimodal models
Video grounding improves with lightweight VLM adapters
Long-context multimodal reasoning benchmarks
Evaluation traps in retrieval-augmented generation
Paperzilla is designed to help you discover, verify, and act on relevant research without treating AI summaries as a replacement for the original source.
01
Source links first
Every result points back to the original paper, dataset, repository, or patent record, so you can verify claims and inspect the source directly.
02
Relevance before volume
Paperzilla is built to reduce alert noise by ranking research objects against the specific topic you care about.
03
Summaries as starting points
AI-generated summaries help with triage, but the source record remains the ground truth.
04
Feedback improves the feed
Your interactions and feedback help the feed become more focused over time.
FAQ
Frequently asked questions
Google Scholar alerts tell you when papers match a query. Paperzilla can accept forwarded alerts, combine them with monitored research sources, and filter the results against your specific research topic so you get a focused feed instead of another noisy inbox.
Yes. You can forward Google Scholar-style alerts and other research-alert emails to Paperzilla so they become inputs to your relevance-filtered feed.
Paperzilla currently harvests papers from sources such as arXiv, bioRxiv, medRxiv, ChemRxiv, and ChinaXiv. Paper-related datasets, software repos, and patents fit the same source model, and additional source support can be added on request.
You define the research topic. Paperzilla compares incoming papers, datasets, repositories, and patents against that topic, ranks likely matches, and improves over time using feedback and recommender-style signals.
Paperzilla can deliver results through an in-app research feed, daily or weekly email digests, RSS/Atom feeds, MCP server access, CLI access, and API access as it becomes available.
Yes. Paperzilla is designed to provide relevance-filtered research context that AI agents and automation workflows can consume through MCP and CLI access.
Paperzilla is a continuous research data monitoring system. It watches research sources and alerts, filters incoming papers and related research data by relevance, and delivers a structured feed for humans, teams, and AI agents.
Get started
Turn scattered research data into a focused feed
Create a topic, connect the sources and alerts you care about, and let Paperzilla surface the papers, datasets, repositories, and patents that matter.