Info about Dataminr Flashcards
from Case studies
Instead of using keywords, Dataminr’s artificial intelligence is able to understand that “killing it on the shoot” isn’t an actual risk. The platform processes billions of units of data from more than 1 million public data sources (in earlier case studies they mentioned 200k then 250k and then 800k) such as social media, blogs and web forums, IoT sensors, audio transmissions, and the deep/dark web. Dataminr is currently being used in more than 650 newsrooms.
The source data can be in multiple formats—text, sound, image, and video— and originate from virtually anywhere in the world, in more than 150 languages, 24 hours per day.
26/03/24
Dataminr often delivers initial signals of emerging events faster than traditional news sources, giving organizations crucial time to plan their response during an emergency.
Dataminr is also useful in remote areas of the world where journalists from large news organizations do not typically operate and where local journalists are slow to report on breaking news.
Dataminr and NYC Emergency Management | Disrupt NY 2017
- ShotSpotter is a technology that can sense when a guns been fired alert the authorities so that they can get to the scene quicker
- Ted: we were thrilled for no money to strike a partnership with city of NY that delivered our technology not only to first responders at the Office of Emergency Management but city wide; this partnership is first of its kind in that it’s the entire city of New York spanning from the fire department to the mayor’s office to first responders to the sanitation department, really everyone across the city government; it started out as a free service
- How is it now? Obviously, we have a business model, let’s just say we don’t we don’t make our money in first response
- Ted: Dataminr has built the first technology that can identify breaking information in social media; many other companies detect trends or analytics, we don’t, we detect that first negative information when it first gets published, so ultimately when we alert it is a headline, it is a alarm bell to first responders to look here and investigate via other sources; a social media signal is never going to be definitive it’s just the most performative early tripwire for breaking information and the state of the world today is that information breaks on Twitter that is something that happens across the world, we’ve done studies that show that unexpected events break on Twitter over 90% of the time before any other source, and the average time across all those events is an hour across the world, so yes it can give a split second advantage but we’re talking about something much more internationally systemic and profound
- Twitter and Dataminr collaborated on building a product for news professionals CNN and the likes so ultimately we perfected this alerting a stream of breaking events
26/03/24
- we decided to focus in on our core value proposition which is breaking news alerts, it’s something that is profoundly impactful in the context of the public sector for first response and it’s a product that truly benefits the society
- the founder of campaign zero said that she’s very concerned about her safety and the ways in which she’s been surveilled by the FBI and and other law enforcement organizations; Ted: unequivocally this is something the Dataminr product does not provide
- the prevalence of fake news is s something that we’re all aware of; in 2014 or 2015, Twitter was used to disseminate information about first or initially a chemical spill that never happened in Louisiana then an outbreak of Ebola that never happened in Atlanta and a police shooting in Atlanta that also never happened and these were coordinated campaigns that involved hundreds of different Twitter accounts, so how do you vet what’s real and what’s not when it’s coming from social media
- many people like to point out that social media can be an
incredibly pervasive way to spread misinformation, what’s most interesting today to mine are having leverage this data for a long time is just how powerful a signal social media can be to verify information you have a set of people out there as eyewitnesses seeing and hearing the world publishing everything they see around them - when the AP Twitter account was hacked a few years back, there was a tweet that was sent out that essentially said there’s an explosion at the White House, now when there would have if there would have been an explosion at the White House there would have been a lot of eyewitnesses around the White House that would have pulled out their phone and said there’s an explosion at the White House; this is the perfect example of how the data was incredibly powerful to verify
Dataminr’s Managing Director Jonathan Barrett
- company started in 2009, 600 employees, HQ in NY, expanded into Europe, now Asia-Pacific is a key area, setting up in Melbourne, Invest Victoria partnership and also AI center in Australia
- original problem: social media started coming to the fore, valuable informaiton hidden in huge amounts of data, identify early alerts
- uses multiple datasets, not only social media but also look at professional sites and blogs and machine-generated data, eg geophysical data centers in the ground in earthquake zones, transponders of airlines and shippers
Dec 2019
26/03/24
- annual subscription-based business model
- 3 core lines of business: support 500+ news agencies with breaking news; corporate risk solutions, particularly around security of assets; emergency response solutions for first responders
- was ranked as No 5 on Forbes A-50 list
- Question on data ownership: what Dataminr uses only publicly available data, if data is not in the public domain, it will be outside
of our sphere
Dataminr CEO Ted Bailey on eyeing an IPO in 2023
- 12-year old startup Dataminr announced a new $475 million funding round, taking its valuation to over $4 billion
- the mix of investors in this round was fairly diverse, one of them was Eldridge, a firm that owns the LA Dodgers and makes Sports Investment
- Dataminr customers are in two main buckets: large global multinational corporations and public sector first responders, for example half of the Fortune 50 rely on data miner signals every day and that ranges from AIG to Goldman Sachs to Shell, when an unexpected event happens in the world, we’re able to warn our customers with the earliest signals on breaking events and critical emerging information
- in this type of environment (2021), all entrepreneurs have to take very seriously the hot nature of the public markets but for us, we decided to pursue this capital round for three main reasons: to accelerate the growth of our corporate business line which doubled a revenue growth now three years in a row, to internationalize to extend our footprint out across Europe and APAC, and to expand our AI platform specking; many other options are all definitely on the table for companies like ours, but we really think public market investors are really waking up to just how promising AI and big data companies are so we’re really excited about an IPO, and to be clear an IPO is on the horizon for us, we’re targeting 2023, this capital round is very much a pre-IPO
March 2023
26/03/24
- C3 is a great example, and Palantir, these are companies that have
really unique fundamental qualities, exceptional gross margin
profiles, and great long-term profitability potential; for us in our trajectory investing in the areas I mentioned over the next couple years is going to put us in the best position for an IPO; I’m long-sighted, we’re building a company for decades to come, just rushing to a hot market is not our goal, our goal is an iconic mobile company and for us 2023 just makes the most sense - how would you help potential investors differentiate between you guys C3 and Palantir, why don’t you run into each other and why isn’t this a zero sun game? Our mission is to integrate all available public data signals to create the world’s first real-time information and event discovery platform for businesses and public sector organizations; 24 hours a day 365 days a year we’re discovering events by detecting information in public data
- a company like Palantir takes external data; a company like C3 is an AI solution that’s brought internal to a company
- Dataminr detects trends and information in external data; we discover signals outside the scope of a corporation or public sector enterprise and this is the way we are differentiated in the AI space
Millenium Live | Dataminr, President of government Dana Barnes
- started my private sector career at IBM where I supported the federal government did a lot of work at the Census Bureau, before that I was in the Marine Corps, then I was 12 years at Microsoft, then moved to Palo Alto Networks, which is very large in cybersecurity
- what drew me to Dataminr was a combination of amazing people and that it’s focused on social good programs, it was key for me was that the company really has an ability to help people, the technology really does good at the individual level and also from a government perspective, the main product for public sector is First Alert and it leverages AI to deliver real-time breaking news alerts helping government agencies make critical decisions in a timely manner
- our CEO Ted Bailey started well over a decade ago when he was in college and studied the history of real-time information technologies, and he was fascinated by how these technologies worked, the printing press and the telegram and the radio, and how they all changed the society, that was something he was really interested in and so he started to focus on other technologies
- he founded the data model back in 2009 and he was really focused on creating the world’s first real-time discovery platform that could detect critical you know breaking information from real-time public data, Twitter, Instagram, Tick-tock, he realized we can use this these sources to provide those real-time alerts, since 2009 there’s been an explosion of public digital data that is unparalleled in human history with millions and millions of digital platforms and IoT sensors, so now when events happen we actually can see them and what we really focus on is how do we deliver those events so I don’t think even Ted could have imagined back in 2009 that would be where we are today but through his vision and through his work with engineers, we built a platform that takes all of that data and can turn it into real-time alerts and look what’s real what’s not real, assess it and push it to the people that need to know right away
Jan 2023
26/03/24
- what makes Dataminr different is our ability not to be a single modality (single modality if you only look at images or if you only look at video) we actually are a multi-modal in our approach, we look at all of those things - image, video, sound, text - and because we’ve built our platform around detecting gobal events were able to see a lot more from 500 000 Plus data sources and growing
- we leverage artificial intelligence, it’s tough for a human being to manage all of that data on their own but when you’re leveraging artificial intelligence, when you’re leveraging machine learning, you can do more and we do it by spanning natural language understanding, we leverage computer vision, sound and audio detection anomaly detection, and machine generated datastream, so over the years of research and experimentation we’ve learned that in many cases an event can only be detected accurately as early as possible by using this deep learning that combines all of those modes and that’s the thing that makes us really unique
- my background at Palo Alto networks is key, as the pre-eminent cyber security firms, but Microsoft was so focused on cybersecurity, because in their early years people thought that their platforms were not secure so much, it got so bad that uh Bill Gates actually shut the company down for a couple days and said we’re not doing anything we’re not talking to
any customers we’re going to solve this security thing right now and become the most secure platform in the planet - the adoption of of IoT and what we call industrial Internet of Things has led to an increasingly interconnected world. when we think about the cyber and the physical security, that’s expanded the the attack surface and it’s blurred the clear lines between cyber security and physical security
- let’s think about ransomware and how ransomware has been used to actually impact the physical world, it’s not just a matter of asking
for ransom to access data but if you can’t run your schools and kids can’t go to school that has a real impact on the physical world, dp my experience the cyber world has helped me to understand the bridge between cybersecurity and physical security and Dataminr kind of sits right in the middle of all of that
Millenium Live | Dataminr, President of government Dana Barnes
- Dataminr’s role in cyber security is evolving just as that space is evolving so we’re continuing to grow so we have some technologies that that we leverage but our public sector product First Alert really does provide early detection of cyber threats we engage and pull data from the dark web so we can see what’s happening in hacker communities
- first and foremost in my space the public sector space organizations benefit, those that are dealing with disasters, think about FEMA in the U.S dealing with natural disasters on a regular basis, we just had the hurricane Ian, or the U.N as an example and some of the things that they do around famine and other things globally those are the type of organizations that really can benefit, school systems also benefit, think Sandy Hook I could go on and on
- I mean we do have a commercial business, it is not my specialty but we do have a commercial side and we see news organizations leveraging the technology, corporations are trying to manage brand awareness so there are I would say there is a whole host of customers that can benefit from our capability that we bring to the table
- how do customers operationalize and most effectively utilize the First Alert platform in order to mitigate and manage risks? It’s a partnership between us and the customers, what FEMA is looking for in terms of of natural disasters might be different than what a Ministry of Foreign Affairs is looking for in terms of protecting diplomats overseas and knowing what’s going on, so understanding what data is important, understanding when you need to know those things is key, so you drive a partnership how do you contextualize so that data so that it’s quickly accessible
- in some cases we integrate into common operating picture systems so that everything is in a single screen, at the end of the day it comes down to use cases, how can we leverage that information more consistently without having to go through a thousand sources, speed is the name of the game
- there is a convergence of the physical and the cybersecurity worlds, they are becoming more and more intertwined, you’ll see us as a company continue to evolve so that we can support those areas, really focusing on critical infrastructure, you will see us continue to focus heavily on social good and supporting organizations that really excel in those spaces
- I think you’ll see us continue to add more data sources and be able to look at rich content to provide alerting to governments and organizations that need it, from a cybersecurity perspective as well as in the physical security space, that’s one of our corporate values
Women in Tech at Dataminr
- I love the sense of community and culture that we’ve been able to maintain even as we’ve grown from the 125 employees when I started to over a thousand employees
- Dataminir in 3 words: spontaneous, foresighted and collaborative
- at Dataminir we are laying the foundation for a new industry I just love coming in every day and helping my team work towards it
- I love that we constantly work towards creating a blameless culture and still we try to learn some of our mistakes and move forward
- there are two things that motivate me to come to work today my team and my work, I love my team and I enjoy brainstorming with them and doing pair programming, when two or more Engineers go together to solve a problem
Dec 2022
26/03/24
- I feel very accepted, heard, and pushed to become better every day
- I’m a director of AI research, it’s challenging but rewarding and fun
- we get to collaborate and love working with people from all different kinds of backgrounds
- I love the people and I love the company values
- to be successful in tech, it’s important to stay true to your values so really focus on what you are good at what you like doing and don’t try and sort of fit into some pre-defined stereotype that doesn’t exactly align with what you’re good at and what you like doing
NightVision Live Fireside Chat with NASDAQ: Dataminr CEO Ted Bailey
- Dataminr integrates now nearly a million different public data sources to detect events, threats, critical information long before any other source, first and foremost we gained adoption across the government, from defense organizations to local law enforcement, but more recently we gained traction across the corporate market, where essentially chief security officers were using our product and what we found was that chief information security officers were being pulled into the programs by chief security officers which led us to build a cyber product and thid is why I’m here bringing data Dataminr’s AI platform to the world of cyber
- the coolest word is multimodal fusion AI, it is at the core of Dataminr’s product, our platform performs trillions of daily computations on billions of individual public data units coming from millions of distinct public data sources, and that computation is something that is done by our multi-modal platform and we’ve pioneered a field of AI called multimodal fusion AI
- if you think about the world’s data, you think about many different
formats – text, image, video, sound, sensors - spitting out numerical streams but what multimodal fusion AI can do is not only to take in data sources of all those various different formats and process them but fuse them together into models that allow for the cross correlation across different types of data sources and different formats; what that allows is just faster detection, more accurate detection and ultimately better actionable intelligence for our customers
May 2023
26/03/24
- we made two major announcements; this is our entrance into the cyber world and there were two announcements we made last week which made it official: we launched our cyber product data Dataminr Pulse for cyber risk and we launched the cyber-physical A2I Coalition, which I am personally thrilled about this is a coalition, where we were fortunate to have really some of the best and leading cyber companies join us to tackle the cyber-physical convergence and that is Mendian and that is Service now and that is Clarity and that is Optive, and this is just the beginning of a coalition that we plan to expand and bring in leaders in cyber to essentially as I like to say go from AI to AI to leverage artificial intelligence to produce actionable intelligence for our customers and this coalition and the members are committed to building joint products into really tackling this cyber-physical convergence area for the CSO (Chief Security Officer) in a way that’s never been done before
- the cyber-physical convergence has been such an interesting thing to be a part of, it’s really twofold in the sense that it’s organizational and it’s functional, the chief security officer and the chief information security officer, their apparatuses are coming together physically to essentially respond together in tandem, but it’s also informational, from a threat perspective it is multi-directional, you have physical events which can ultimately ripple on to the cyber world, and you have cyber incidents which can have tremendous and catastrophic impact on the physical world;
- a very famous example of the Colonial pipeline attack, which really helps illustrate cyber-physical convergence, here you had a ransomware attack shut down a major pipeline across the east coast and ultimately what we were able to detect were the many physical world ripple effects of that you know ultimately, the operators were sensing the fuel disruptions and starting to post these messages online and then ultimately the ripple effects, like shortages at the airports, flights being canceled
- but it can go the other way, for example when Ukraine was invaded by Russia, the moment that happened that first indication that ultimately you know the rush across the border was a very valuable signal to the chief information security officer to go Shields up, you never know when geopolitical crisis can come concurrent with cyber activity in cyber warfare, but it doesn’t have to be something as profound as you know a nation-state conflict, even a small localized fire could be happening that’s near an IT facility, or that’s near an OT asset or or some IoT infrastructure that also could pose an extremely significant cyber risk so really cyber physical convergence is multi-directional and our capability tries to tackle that from both sides
- a cyber-physical attack has a life cycle and it starts with a vulnerability and then that vulnerability is exploited, what we’ve found in the data sources that we’ve been able to take in is that detecting vulnerabilities early and prioritizing them and ultimately finding trending vulnerabilities is something that our AI platform can do very well, for example we were able to detect log 4J early when it first got out, when there were first researchers seeing the vulnerability and that’s a really powerful example of how out there on the external surface of the internet you can get some of the fastest, most actionable intelligence about vulnerabilities
- Dataminr is a company which you know intends to go public and we are excited to continue to grow our business both the government and the corporate market, coming out of the government space and the physical security space and now the cyber space we diversified our portfolio, we continue to expand across the corporate market from business intelligence to supply chain to insurance, frankly what we’ve found is when it comes to detecting The World At Large and what’s happening and physical incidents, believe it or not there are a lot of use cases, so we continue to allow our platform to power a number of different use cases and verticals
Interview of Ted Bailey by CNN International on Covid
- Dataminr found eye witnesses’ accounts in Wuhan in Chinese, photos of officials disinfecting public areas
- they detected it a week before the CDC report
- also detected outbreaks in Iran, SK, Italy, Spain
- they found that eye-witness reports, combined with AI, can create what they call “ground truth”, an accurate on-the-ground understanding
- the key is how to apply the computing power of AI
- we detect billions of public data units per day across 10k public data sets in 150 languages and perform trillions of computations to find that needle in a haystack
- there is a lot of buzz about “outbreak science”, data miners always specialisein this, not in terms of a disease, but in the context of detecting early information when it first emerges, long before it goes viral
2020
26/03/24
from Google
Bailey founded Dataminr in a partnership with two other friends, San Hendel and Jeff Kinsey, who also attended Yale Univ
When we dug a little deeper to learn more about who inspired Ted Bailey, we learned that the late actor Robin Williams had an impact on him. He pointed out that Williams in his various roles was a great many things, but above all the rest, the man was one of a kind. This information was taken from a quote made by former President Obama and it resonated with Bailey and inspired him to be that one-of-a-kind guy.
Dataminr CEO Ted Bailey on CNBC at Davos 2016
- raised $180m in funding
- backed by Venrock, Institutional Venture Partners, Fidelity, Wellington and Credit Suisse
- Merck did not put a press-release about a free antidote to Ebola, recent cases in Sierra-Leone, but tweeted it
- Twitter is a global sensor network
Jan 2016
27/03/24