Recorded Future – Harnessing intelligence at the scale and speed of the Web
I am at the Recorded Future's first user conference down in Washington today, watching Christopher Ahlberg and guests takes us through the next generation in business intelligence. Team Atlas has a "Big Focus on Big Data" and this is one of those days where I'm comforted that we're on to some strong opportunities in this field.
At Atlas my predecessors (and Philippe Claude in particular) played an important role in funding the first generation of business intelligence success stories, including for example Business Objects, iLog or Spotfire (Christopher's previous success). In our latest fund, we've continued with this focus and have a number of companies focused on providing intelligence under the suitably vague moniker of Big Data (Dataxu, Adsafe Media, Recorded Future, Hopper Travel and others).
In the past business intelligence was enterprise centric and non real time. It was focused on understanding history, whether in the form of Entreprise Search (Endeca, Autonomy) or traditional Business Intelligence (BO, Cognos, Tableau, Qliktech). Where understanding trends across time and mapping the future was concerned, the only plays out there dealt with highly structured data, usually based on mathematical approaches (MatLab, SPSS, SAS).
So conceptually Recorded Future is about a simple concept : horizon scanning at the scale of the web. If knowledge management is about analyzing history and trying to derive rules for the future, Recorded Future is about connecting the dots of a fast moving present and uncertain future.
As I'm listening to Recorded Future users from the intelligence community and large corporate users, I'm struck by how far we've come since my days on the board of Xerox-PARC spinoff Inxight. We've got a perfect storm of technologies coming together that makes a company like Recorded Future possible.
Next Gen Intelligence Infrastructure
A company like Recorded Future is hard to build today (just ask Quid), but would have been a pipe dream 5 years ago. We've got a number of connected dots that help make a company like this possible.
Rich Data Substrates. Yes, thank you Google for indexing every document out there. Thank you Facebook for the Open Graph. Etc, etc. Data wants to be free. Before you were linking 60 instances of SAP to build your BI cubes, now we're throwing the web at you on top.
Real Time Data. The real game changer is the need / ability to harnessing real time data. If you're trying to predict an uprising in Syria, newspaper scanning isn't going to help you. If you're trying to map an information leakage problem, you'll need to track how a leak travels across the social graphs. Real-time is the real game changer in providing actionable intelligence.
Big Data Infrastructure & Big Data Analytics: I think of Big Data in a simple way. We used to have predictive algorithms that would crunch large data sets, now we have huge data sets and more importantly complex interactions that need to be observed instead of theoretically modeled ex ante. Intelligence is derived from the actual data instead of making assumptions about models that will accurately describe the world. Anyways, we know have both the databases and the analytics to allow us to harness data at the scale of the web. Our newest partner Chris Lynch, ex Vertica CEO, knows a thing or two about that.
The Power of Context
Context is the great enabler of Big Data. It's what helps separate signal from noise in real time environments. Whether you're extending your current service through a new mobile app or trying to derive intelligence from a pile of data like Recorded Future, context is what's going to drive the appropriate form of user interaction and help you harness your sea of data in a relevant way.
If you're mixing internal intelligence with competitive data and market data, you're now able to derive "contextual intelligence" from these data mashups enhanced by powerful analytcs and rich visualizations. To put it more simply, the world of business intelligence used to be about loading internal data into data warehouses and running scripted analytics against this. Now, we're looking at context rich, personalized, real time analytics at the scale of the web.
Impactful use Cases
The revolution is either enabling or fueled by new use cases. Here's a few of things you can do today with a platform like Recorded Future:
Providing physical security: if you're a large manufacturer, can you predict which production facility or port might be affected by unrest ? Can you protect your C-Level execs are they travel across the world ? If you're at the CDC, can you map swine flu outbreaks in real time across the globe ? Can you identify leading indicators of a likely bird flu outbreak from a government that has a tendency to suppress data ?
Catching Baddies: can you predict where senior Al-Qaeda in Maghreb operatives are moving ? Can you identify the real media influencers that drive the al Qaeda messages ? Can you trust that new informant or is he a double agent ? Can you dynamically geo-locate information that relates to your top targets ?
Managing information leakage: what if you are a leading consumer electronics company worried about product releases leakage. When a leak occurs, can you map the ways in which the leak occurred and spread across news and social networks ? Can you track back the movement of your employees and correlate them with the movement of the blogger or journalist at the source of the leak?
Map cyberthreats: when worm variants start spreading faster than Symantec can push out the relevant defenses, can you map in real time the spread of the infections and push defenses in the right way ? Can I determine whether a threat is accelerating or slowing down ? Is there an identified threat trending across the hacker community ? Etc.
Map competition: If you're trying to get a view for what a competitor is doing and you're say, a large FMCG company, it's tough to get a complete view of your opponents' strategy. But what if you could map joint ventures, new commercial agreements on sourcing and distribution, patent activity and investment activity across the globe and across time ? What if once you get a whiff of a possible acquisition you can quickly analyze the impact on your position in each market, whether there are regulatory levers you can pull in specific markets to block a merger, and so forth ?
Use cases we cannot talk about : we got a bunch of those. Makes for interesting pipeline reviews at the board meeting: Agency 3, Product X. You get the picture. Terrorism prevention, nuclear safety, covert ops, whatever these guys do.
Nailing the elusive billion dollar opportunity
You can clearly see the challenges of building a company like Recorded Future : it has everything to do with nailing repeatable and highly productizable use cases that you can use to get the company to scale effectively. With a backend that works (i.e. Technology that's proven), we'e faced with an embarrassment of riches as to where to take the products that sit on top of the FR platform and help us scale revenues fast and profitably.
It's a playbook that's been written many times before. The beauty of startups of course is that every time is different. With an opportunity of that magnitude the founding team has been working hard on getting the best assets on side. That includes building a team of senior folks out of Spotfire, Endeca, Google, Vertica and elsewhere, as well as attracting strong capital backers early, including Google Ventures (Rich Miner), IA Capital Ventures (Roger Ehrenberger) most recently Business Objects founder and CEO Bernard Liautaud, now a VC at Balderton.
We've got the tech, the people, the capital and the reference clients; watch this space for Boston's next big enterprise success.