From raw signals,
structured foresight.
Build knowledge graphs from documents. Run multi-agent simulations. Generate source-cited intelligence reports with confidence scoring.
From documents and scenarios
to simulation-backed intelligence.
Signal Processing
Ingests documents, extracts entities and relationships, and constructs a structured knowledge graph. Raw information becomes queryable intelligence.
Scenario Simulation
Deploys LLM-generated agent personas into synthetic social environments. Multi-agent dynamics model how events, opinions, and decisions could unfold.
Structured Foresight
Generates source-cited intelligence reports with confidence scoring. Every claim traces back to evidence — graph data, web sources, or simulation outcomes.
Four-stage pipeline.
Each step is discrete and inspectable.
Graph Build
Upload documents — PDF, Markdown, or text. Phoring parses content, extracts entities and relationships via LLM-driven ontology generation, and constructs a knowledge graph in Zep.
Agent Setup
LLM-generated agent profiles emerge from graph entities. Each carries a persona, stance, behavioral parameters, and platform-specific interaction models tuned to the scenario.
Simulation
OASIS deploys agents into synthetic Twitter and Reddit environments. Actions, reactions, and emergent discourse unfold across configurable rounds — streamed in real time.
Intelligence Report
The Report Agent synthesizes graph data, web intelligence, and simulation outcomes into a structured document — source-cited with inline references and per-section confidence scoring.
Built for structured thinking
about uncertain futures.
Policy Impact Analysis
Model how regulatory changes propagate through public discourse. Simulate stakeholder reactions across platforms before policy announcements.
Market Reaction Forecasting
Feed financial signals and market context into the pipeline. Surface how trader sentiment and media narratives could evolve after key events.
Crisis Response Modeling
Upload crisis documentation and simulate rapid-response scenarios. Identify which narratives gain traction and where sentiment fractures.
Technology Adoption
Map the stakeholder landscape around emerging technologies. Simulate adoption patterns, resistance points, and discourse evolution over time.
Geopolitical Scenarios
Build entity graphs from geopolitical briefs. Run scenario simulations across multiple possible outcomes with confidence-weighted forecasts.
Public Discourse Simulation
Understand how information spreads through social platforms. Model opinion formation, polarization dynamics, and consensus pathways.
Different by architecture,
not by marketing.
Not a chatbot.
A simulation engine.
Phoring doesn't generate answers from a single model prompt. It runs structured multi-agent simulations where synthetic personas interact, debate, and surface emergent patterns across configurable rounds.
Not prediction from thin air.
Grounded in evidence.
Every forecast traces back to document-sourced entities, web-retrieved intelligence, and simulation-observed dynamics. The quality of the input directly shapes the output.
Not a black box.
Source-cited and scored.
Reports include inline references [1][2][3] to specific sources. Each section carries a confidence tag — HIGH, MEDIUM, or LOW — based on how many independent data points support it.
Not a single perspective.
Multi-agent, multi-model.
Optional multi-model consensus validation runs independent LLMs as validators — scoring on coherence, precedent, and risk before producing a consensus summary appended to the report.
Transparent by design.
Every step is auditable.
LLM-driven ontology extraction → entity and relationship mapping → Zep Cloud storage → downstream querying for agent profiles, simulation context, and report generation.
Entity-scoped queries via Serper and NewsAPI. Articles scraped and processed up to 4,000 characters each. Social content sourced via Google Search indexing.
OASIS framework spawns synthetic environments where LLM-generated agents interact based on assigned personas, stances, and behavioral parameters. Results stream in real time.
ReACT loop pulls from knowledge graph, web intelligence, and simulation data. Claims are backed by inline numbered references with a full sources section.
Each report section tagged [HIGH], [MEDIUM], or [LOW] based on independent tool-sourced data points. Reflects evidence density — not a guarantee of accuracy.
Optional. Up to 3 independent LLM validators score predictions on coherence, precedent, and risk. A consensus summary is appended to the final report.
Upload documents,
define your scenario.
Drop your source files and describe the scenario you want to simulate. Phoring handles the rest.
Drop files here or click to browse
PDF · MD · TXT — up to 50 MB
Start mapping
what's ahead.
Upload your documents, define a scenario, and let Phoring build the intelligence.