Rolli
For researchers & policy analysts
94.2% detection precision on known CIB datasets

Your dataset counts social activity. Ours tells you what's real.

Rolli provides citation-grade authenticity scoring, reproducible exports, and methodology documentation written for methods sections — so your findings hold up.

8 platforms, 1 schema Authenticity metadata Reproducible exports Longitudinal tracking
Research Infrastructure
200M+
data points indexed
40+
languages supported
8
platforms unified
Real-time
data freshness

Aligned with research frameworks from

Stanford IO·Columbia Journalism Review·Harvard Shorenstein Center·DFRLab

Trusted by communications, security, and research teams

94.2%
authenticity scoring precision
Rolli IQ detection precision on known CIB datasets
30–40 hrs
saved per research project
vs. manual data collection and normalization
13%
avg authenticity on coordinated campaigns
Coordinated inauthentic behavior benchmark
Dataset-grade outputs
6 capabilities

Built for methods sections

Every Rolli dataset includes authenticity scoring, reproducible query exports, and structured intelligence designed to hold up under peer review.

Reproducible Collection
  • Boolean AND/OR/NOT queries — fully reproducible
  • 8 platforms, standardized versioned schemas
  • Timestamps, post IDs, author handles for citation
Sentiment & Emotion
  • Post-level sentiment + emotion classification
  • sentiment_target and sentiment_reason fields included
  • Methodology docs ready for your methods section
Topic Trees
  • Structured conversation branches from raw thread data
  • Branch summaries + cluster-level sentiment
  • Exportable for qualitative or mixed-methods coding
Coordination & Authenticity
  • Coordination score 0–100 per narrative or account cluster
  • Frequent-poster and distribution metrics
  • Deepest signals on X; cross-platform available
Platform Summaries
  • Per-platform sentiment classification
  • Per-platform narrative summary across time periods
  • Built for longitudinal tracking and comparative analysis
Built for Reproducibility
  • Webhook + batch exports for long-term datasets
  • Usage endpoint for audit trails and replication
  • Methodology documentation available on request
View methodology documentation →
Live Data Preview

The data your methods section will reference

Authenticity metadata, platform coverage flags, and coordinated-activity signals — structured for reproducibility and peer review from day one.

Capabilities

Data that survives peer review

Citation-grade authenticity scoring, reproducible exports, and methodology documentation written for methods sections — so your findings hold up.

Cross-platform signal data

Normalized data from 8 platforms — X, Reddit, YouTube, Facebook, Instagram, Threads, Bluesky, and LinkedIn — in a single schema. No platform-by-platform collection scripts, no ad hoc normalization work. Rolli handles the pipeline so you can focus on analysis.

Unified cross-platform schema
Platform coverage flags per record
Consistent normalized timestamps
8 sources, 1 export format

Reproducible exports

Structured JSON with documented schema versions, platform_coverage flags, and field-level provenance metadata. Every export can be fully described in a methods section.

Documented schema versions for citation
Field-level provenance on every record
platform_coverage flags for source completeness

Longitudinal narrative tracking

Track how a narrative evolves across time and platforms. Configure retention windows appropriate for your study timeline and receive consistent, timestamped data snapshots.

Configurable retention windows
Consistent timestamped snapshots
Study-period data integrity
Coordinated Behavior Scan
● LIVE
3 clusters detected

Authenticity metadata for methodology

Every item includes authenticity_score, coordinated_flag, and velocity_score — fields that enable methodological controls for inorganic amplification in your study design.

authenticity_score0–100 organic-to-coordinated range
coordinated_flagboolean CIB signal per record
velocity_scoreamplification rate vs. organic baseline
Peer-Reviewable Output

Built to survive academic scrutiny.

Every Rolli export is structured for citation, methods documentation, and reproducibility — so your findings hold up at the journal level.

Citation-Ready Data

Schema version and collection timestamp on every item — everything a reviewer needs to verify your source.

Structured Exports

JSON and CSV outputs with field-level provenance metadata — drop directly into your analysis pipeline.

Methods Documentation

Methodology documents written for methods sections — ready to paste into your paper without modification.

Reproducible Configs

Study configurations your co-authors can replay — enabling multi-researcher replication without ambiguity.

Who Uses Rolli

Four research communities that depend on knowing what's organic

From university labs to policy think tanks — one truth layer built for academic rigor.

Academic researchers

Social scientists, communications scholars, and computational researchers studying narrative diffusion, political communication, and online discourse at scale.

Request access →

Policy analysts

Government-adjacent analysts and non-profit policy researchers tracking legislative narrative environments, public opinion signals, and influence operations affecting policy outcomes.

Request access →

Think tanks

Research organizations producing reports on media ecosystems, platform governance, and information integrity that require cross-platform, citable data sources.

Request a demo →

AI safety researchers

Teams studying LLM-generated narrative amplification, AI-assisted influence operations, and the detection of machine-generated content at social scale.

Talk to an engineer →

Get your first narrative intelligence brief in 24 hours

Talk to a researcher who built this. No sales cycle required.

Methodology Readiness

What your methods reviewer will ask — and what Rolli gives you to answer

Question 1 of 3

How can I trust the data quality?

Our Answer

Every item in the Rolli schema includes a platform_coverage field indicating which fields are populated for that source, and a data_quality_flags array for known collection anomalies. We document known limitations rather than hiding them.

Question 2 of 3

Can I describe your methodology in a paper?

Our Answer

Yes. Research users receive access to methodology documentation that can be referenced in methods sections. We are committed to methodological transparency and will work with you to ensure your description is accurate.

Methodology documents include signal weighting, versioning, and field-level provenance — written to satisfy peer review requirements at journals requiring full methodological disclosure.

Question 3 of 3

Our budget is constrained. Is there a path forward?

Our Answer

Yes — and often at no cost. Qualifying academic institutions and non-profit policy research organizations receive sponsored access or significant discounts (typically 40–60% off). Request API access with your institutional affiliation and research purpose. We review within 2 business days.

Request API Access

Next onboarding cohort: 14 spots remaining · Setup in under 10 minutes

Research Brief Preview

How academic researchers use Rolli

A sample query output — showing the structured, reproducible format built for methods sections and peer review.

rolli-api · narrative-query · research-export
CITATION-READY
Sample Query
GET /v1/narratives?query="AI election interference"&window=30d&platforms=all&include_authenticity=true
Results Format (abridged)
"authenticity_score": 0.24
"coordinated_flag": true
"velocity_score": 8.4
"platform_coverage": [x, reddit, bluesky, threads]
"methodology_version": "rolli-iq-v3.0"
"data_quality_flags": []
Export as:CSVJSONAcademic CitationMethods section ready

All fields documented in methodology guide · Versioned schema · Field-level provenance metadata included

FAQ

Methodology, data quality, and access

Academic & non-profit pricing available.Contact us with your institution →
600+

narrative investigations analyzed since 2020

Intelligence your research can cite. Data your leadership can trust.

Institutional review within 2 business days