Proprietary ML technology enabling automated features for accurate tracking and insights across an array of sources and social media platforms
Track and analyze world leaders, politicians, companies, non-state actors, and events
Go beyond basic Boolean keyword searches with NLP
Always on community listening and tracking
LASTING DATA RETENTION
EASY-TO-USE, REAL-TIME DATA EXPLORER
Consolidate multi-source intelligence streams into unified displays, enabling analysts to monitor cross-domain patterns.
Surface trending hashtags, influential creators, and real-time executive summaries side-by-side.
GenAI-powered tool to generate structured analysis instantly.
Produces briefing-ready intelligence for decision-makers.
Create alerts with precise, analyst-defined conditions and flag high-risk content.
Alerts appear in a dedicated feed, ensuring critical developments are never missed.
Identify the regional origins of creators, offering governments context for understanding media ecosystems and geopolitical dynamics.
Attribution powered by multiple signals:
Pendulum is a dual-use platform built on proprietary machine learning to help government agencies track and understand information flows across today’s fast-moving media environment. In a world where stories can leap from obscure forums to front-page headlines within hours, traditional monitoring tools miss critical context. Pendulum ingests PAI video, audio, text, and imagery content from more than 25 platforms in over 75 languages to capture the signals shaping narratives, public sentiment, and policy debates. Backed by leading investors and informed by deep national security experience, Pendulum is built to meet the mission needs of government analysts.
Multi-Modal Information Collection
Pendulum continuously analyzes millions of data points daily across text, audio, and visual content, revealing how strategic concepts, emerging threats, and policy narratives are interpreted, contested, and manipulated across communities. Advanced language models generate Boolean extensions and semantic queries that power discovery across news outlets, podcasts, and major platforms such as YouTube, TikTok, Reddit, Instagram, Facebook, RedNote, and Telegram. Multilingual capabilities—spanning search seeding, automatic transcription, and translation—enable agencies to capture critical conversations that originate outside English-language media. For audio and video, AI models automatically detect, transcribe, and extract relevant segments, streamlining analysis of vast libraries of content without losing context.
Information Detection & Analysis
Comprehensive filtering tools allow analysts to narrow focus by timeframe, platform, and content type. Integrated OCR reveals hidden text embedded in memes and videos, while sentiment analysis highlights patterns such as toxicity and threatening language. A continuously updated snippets feed provides direct visibility into the underlying data driving media insights, making it easy to dig deeper into the most relevant content. Analysts can also add creators of interest to monitoring lists, enabling teams to track influential or inflammatory voices as narratives evolve across platforms. Looking ahead, geolocation capabilities are in development to surface deeper insight into the origins of these creators.
Automated Reporting and Alerts
AI-generated reports and executive summaries distill complex data streams into actionable intelligence, with direct links back to original content for deeper investigation. Reports are fully customizable in scope, analytical depth, and frequency—providing curated visibility into operational priorities and emerging risks. Smart alerts notify analysts when user-defined conditions are met and display results in a dedicated feed, ensuring critical developments are never missed. All insights can be consolidated into a customizable dashboard that unifies data streams across multiple topics of investigation, enabling side-by-side comparison and a comprehensive view of the information environment.