Dataminr
From AltData.wiki, The Alternative Data Encyclopedia
Dataminr is a private American artificial intelligence company that detects breaking events, threats and risks in real time from public data sources. Its platform analyzes signals from more than one million public data sources using proprietary large language models to deliver early warnings to corporate security teams, government agencies, newsrooms and financial institutions.[1][5].
Founded in 2009 and headquartered in New York City, the company is known for products including Dataminr Pulse for corporate security, First Alert for the public sector, Dataminr for News and Dataminr for Cyber Defense.[1] It came to wider public notice when an alert about the death of Osama bin Laden circulated 23 minutes faster than major news organizations reported it.[2]
What It Does
Dataminr's AI platform provides the earliest warnings of high-impact events across public datasets, converting raw digital chatter into machine-readable alerts with context for response.[1][2] In the corporate market its Pulse product supports crisis response with playbooks, messaging tools and post-event documentation, while First Alert serves first responders and public sector agencies monitoring natural disasters and other emergencies.[2]
Data And Methodology
The company states that its discovery layer draws on more than one million public data sources and more than 50 proprietary LLMs, combining what it calls Multi-Modal Fusion AI for event discovery, generative AI for description and agentic AI for context.[1] The platform's origins lie in real-time detection of emerging information on social media, which was extended into broader event detection after Twitter restricted data access in subsequent years.[2]
Products
Dataminr's product line comprises Dataminr Pulse for Corporate Security, Dataminr Pulse for Cyber Risk, First Alert for public sector situational awareness, Dataminr for News used by journalists in more than 1,500 newsrooms, and Dataminr for Cyber Defense launched in late 2025 following the announced acquisition of ThreatConnect.[1] The company also operates a developer portal for integrations.[1]
Buyers And Use Cases
According to the company, its clients include more than 100 U.S. government agencies, over 20 international governments, two-thirds of the Fortune 50 and half of the Fortune 100.[1] Documented use cases span emergency management, executive protection, travel risk, supply chain monitoring, insurance and financial services decision-making, as well as humanitarian work through a partnership with UN Global Pulse begun in 2019 to equip UN personnel with crisis alerts.[1][2]
History
Dataminr was founded in 2009 by Yale graduates Ted Bailey, Sam Hendel and Jeff Kinsey.[2] In 2014 it partnered with CNN and Twitter to launch Dataminr for News, a tool alerting journalists to emerging information on Twitter in real time.[2] Its first acquisitions were UK geovisualization platform WatchKeeper in July 2021[4] and Copenhagen-based crisis response platform Krizo in October 2021.[2] In October 2025 it announced the acquisition of cyber threat intelligence firm ThreatConnect for $290 million in cash and equity.[3]
Landscape
Dataminr competes in the real-time event detection and geopolitical risk segment alongside firms such as RavenPack, Signal AI, Causality Link and Heckyl, which press coverage has grouped together as vendors moving geopolitical risk data into fund management.[2] Within alternative data portfolios it is typically classified as a news and event intelligence provider rather than a structured dataset vendor.
Compliance And Controversies
Investigative reporting by The Intercept has raised questions about the use of Dataminr alerts by police departments, including monitoring connected to the 2020 George Floyd protests, U.S. Marshals surveillance of abortion rights protests reported in 2023, and tipping of the LAPD to pro-Palestinian demonstrations between October 2023 and April 2024 according to a 2025 report.[2] The company has stated that First Alert identifies breaking news events without regard to demographic characteristics of posters and offers no geospatial analysis features.[2]