2015 conferences
Biotech and Beyond

Making Stuff

Innundata and the Fog of IT
Regional Meetings
Prior conferences
Upcoming conference
more about us



Innundata and the Fog of IT
July 21–22, 2015
Philadelphia, PA

Library Selection
Dataclysm: Who We Are (When We Think No One's Looking) (Crown, 2014)
By Christian Rudder


6:00 PM
First-Timers Reception — Independence Room
6:30 PM
Welcome Reception — Grand Ballroom Foyer
7:00 PM
Dinner — Georgian Room

7:30 AM

Breakfast — Georgian Room

8:30 AM
Conference Welcome - Grand Ballroom
Len Kleinrock, TTI/Vanguard Advisory Board
8:45 AM
Big Data Tools for the Dark Web
Christopher White, Principal Researcher, Microsoft, Formerly Program Manager, DARPA
DARPA has been investing in technology for data science and building open-source tools for applications ranging from counter threat finance, through radar operations and cancer research, to anti-human trafficking. This presentation will cover recent work at DARPA, experience building real-world applications for defense and law enforcement to analyze data, and the future of computer science as an enabler for content discovery, information extraction, relevance determination, and information visualization. The talk will be a mix of background, examples, and software demonstration.
9:25 AM
Computer Vision, Deep Learning, and Automated Image Captioning
Larry Zitnick, Principal Researcher, Microsoft Research
Since the 1960s, artificial intelligence researchers have chased the elusive goal of having a computer describe a photo using natural language. In just the last year, after a flurry of activity by top universities and industrial research labs, significant progress has been made. Today, Microsoft, Google, and others are able to suggest captions for photos that are commonly indistinguishable from those written by humans. What enabled the field to move so quickly? Concurrent advances in big data, deep learning, and the use of recurrent neural networks borrowed from the machine translation community. Still, current approaches have limitations. We’ll discuss them, as well as the future directions of the field.
10:05 AM
Coffee Break — Ballroom Foyer
10:35 AM
A Few Lessons I’ve Learned
Dan Wolf, Chief Executive Officer, Cyber Pack Ventures
Cybersecurity, INFOSEC, Information Assurance—regardless of what you call it, it’s the ‘wicked problem’ of our time. And protecting yourself and your business has never been more challenging. Wolf spent almost 40 years working on intelligence activities and protecting U.S. government secrets for the National Security Agency. He also teaches graduate students how to think about the security challenges they will face, in classes that are grounded in practical solutions. In this presentation, he shares a few lessons learned and why he is hopeful for the future.
11:15 AM
Explanatory Analytics to Drive Revenue
Ari Tuchman, Co-founder and Chief Executive Officer, Quantifind
People tweet and post on social media about new movies all the time, yet Hollywood cannot translate that data into increased ticket sales. A major telecom company wins awards for its advertisements, but those ads don’t reduce customer churn. In both cases, social media contains key clues for them to meet their goals. We’ll look at these and other examples of using signal processing methods and data analytics to look at social media chatter and predict buying behavior.
11:40 AM
Measuring the U.S. Economy in Real Time
Micheline Casey, Chief Data Officer, Federal Reserve Board of Governors
The world is a very different place from that of 1922, the year the Federal Reserve began producing monthly statistical forecasts of Industrial Production. To be sure, the Fed now generates other, more timely Fed measures of activity. But today’s U.S. economy is larger than ever, in scope, scale, and complexity, and, since the recent financial crises, the financial regulatory ecosystem has also changed. At the same time, new tools are available to economic forecasters, brought about by rapid advancements in technology, business innovation, and the consequent availability of far more real-time data than has ever previously existed—a massive torrent of new complex data that is both required by the new regulatory regime and available via third parties and open-source mechanisms. All this demands new ways of thinking, and best practices, as they are being applied in many other sectors.
12:20 PM
Members’ Working Lunch & Lunch Session — Georgian Room
1:35 PM
Ethereum: Building the Decentralized World, Made Easy
Vitalik Buterin, Founder, Ethereum
Ethereum is a community-driven project aiming to decentralize the Internet and return it to its democratic roots. It is a platform for building and running applications that do not need to rely on trust and cannot be controlled by any central authority. The platform makes it possible for any developer to build and publish next-generation distributed applications. Contracts are the main building blocks: Sending a transaction to a contract causes its code to execute. Contracts can store data, send transactions and interact with other contracts; they can be used to build currencies, financial derivatives, voting systems, decentralized organizations, data feeds, title registries and thousands of other applications. Ethereum can be used to codify, decentralize, secure and trade almost anything.
2:15 PM
Data-Driven Trend Spotting
Luca Morena, Co-Founder and Chief Executive Officer, iCoolhunt SpA/Nextatlas
Einstein once said, “Not everything that can be counted counts, and not everything that counts can be counted.” Nonetheless, data now pervades business: even HR is using it for workforce planning; even salespeople are using it to close deals. The next frontiers for data-driven decision-making are the creative professions. Rather than interfere with the creative process, we will see a new capability that we might call “data-driven inspiration,” which taps into the vast reservoirs of creativity and intelligence to be found on social networks, and a new breed of “creative quants,” who will emerge to exploit them.
2:40 PM
Information Security Without Hardware or Software
Ryan Lackey, Principal, Security Practice, CloudFlare
Some of the same web data analytics that provide insight into your visitors and users can also look for online attacks. Indeed, for most websites, threats and crawlers make up 20% to 50% of traffic. It’s traffic every website should understand, but most analytics services ignore. A large service provider will, as well, see enough global web traffic to share threat information with the rest of its community.
3:20 PM
Coffee Break — Ballroom Foyer
3:50 PM
Pixxcell: Every Image, Everywhere
Melissa Mercer, Founder, Pixxcell
There is no central platform for the worlds images. They are spread across millions of different websites, thousands of museums, libraries and archives, or stored in boxes and attics. Our entire history is documented yet no one has organized it. In 100 years time how are we going to look back and see our past? 
4:25 PM
Improving City Life Through Data
Ben Wellington, Quantitative Analyst, Two Sigma
Every data set that the city releases tells a story, and stories can lead to policy changes. For example, once we identify the worst fire hydrant in New York City from the standpoint of parking tickets ($55,000 annually), officials just might repaint the roadway markings. When we discover why half the taxicabs in the city collect more in tips than the other half, and what hidden credit card calculations cause that, we can reconcile the algorithms. Hacking city data is fun, and can lead to real change.
5:00 PM
Close of First Day
6:00 PM
6:15 PM
7:15 PM

Buses to City Tavern, 138 S 2nd Street

7:30 AM

Breakfast — Georgian Room

8:30 AM
Data in Motion: Cities, Logistics, and the NBA
Rajiv Maheswaran, Chief Executive Officer, Second Spectrum, and Research Assistant Professor, Computer Science Department, University of Southern California (@RajivMaheswaran)
There has been an explosion in data about spatiotemporal information: the locations of various entities and how they change over time. This includes logistics data in corporations and the militaries, tracking vehicles and people in cities, and recently and prominently in professional sports. The main problem with spatiotemporal data as is with other data sources is the ability to extract meaningful information about it to enable intelligent modeling and simulation. We have started in sports to build the “science of moving dots” and this has enhanced the way people coach, play, and watch sports and shows the path of how we can improve the activities associated with other trackable entities.
9:10 AM
Stories That Data Tells
Shonali Krishnaswamy, Head, Data Analytics Dept., Institute for Infocomm Research, Singapore
This talk will focus on three real-world cases studies and demonstrations that show the opportunities for drawing rich insights from “not so” rich data. The first case study will be on the topic of Predictive Audit where we show how Data Analytics has led to a transformational impact in a large financial institution by enabling Audit to preventive/pre-emptive of future risks and irregularities, rather than reactive/responsive to past risks and irregularities. The second case study will focus on “Tell Tale Trajectories,” showing how “not so” rich and sometimes rather sparse data such as public transport cards and mobile cell tower connections can be leveraged to draw deep insights about users, their preferences, lifestyles, and behavior. The third case study will focus on predicting faults and failures and show how to overcome the challenges of making accurate predictions when there are very few relevant data points/exemplars to learn from. The talk will also briefly examine the emerging world of Big Data Partnerships in sectors such as Finance, Insurance, Telco, and Health/Wellness.
9:50 AM
High-Speed Adaptive Stream Processing and Analytics
Nagui Halim, IBM Fellow and Director, Chief Architect, Big Data
Big data and analytics solutions are dominant subjects in the IT world right now, and it’s easy to see why. By detecting and acting on previously unseen trends, patterns and information of all kinds — in areas ranging from customer demand to infrastructure performance—organizations can create many compelling forms of new value. A new tool, stream processing, can empower organizations to get insights from data far more quickly, and in far more ways, than ever before—even in cases where insights are needed in real time, or very close to it. Stream processing represents a fundamental break from the traditional paradigm. Instead of collecting and storing the data first, which creates a significant delay, you run analytics against the data as it becomes initially available to the organization. Stream processing is already being in a wide range of use cases, including stock trading, network/infrastructure performance, social media data, government and law enforcement, energy, and security.
10:30 AM
Coffee Break — Ballroom Foyer
11:00 AM
Data Wrangling
Joe Hellerstein, Co-Founder and Chief Strategy Officer, Trifacta
Data used to be a problem managed by IT. Increasingly it’s an opportunity being exploited outside IT—by the people who understand how the data can improve business. Unfortunately, working with raw data was never easy, and technical skills vary widely outside IT. How can those gaps be bridged? Via innovations in the way we work with data: new technologies and interface designs that marry software intelligence with novel user experiences to radically simplify the way that people can do their own data wrangling.
11:40 AM
Organic and Artificial Perception: Discovering Meaning in Visual Data 
Ahna Girshick, Head of Product & Partnerships, Enlitic
While our brains have built-in high-resolution sensors, they can’t parse CSV, JSON, or SQL. Fortunately, when data is turned into images, we can easily understand its patterns and meaning—in fact, our brains evolved to optimally apprehend the visual data in our world. Understanding that process has lessons for both neurophysiologists and computer researchers, but particularly for technologists and designers looking to find and enhance meaning in data. Key advances have been made recently by new vision algorithms (so-called deep learning) that have had unparalleled success in automatically discovering meaning in visual data, yet they are still subject to some of the same idiosyncrasies of our brains. Examples can be drawn not just from data visualization but also cinema, advertising, typography, and even medicine.
12:20 PM
Designing and Building Data Visualizations
Wes Grubbs, Founder, Creative and Technical Director, Pitch Interactive
Data visualization is a practice of seeing and crafting patterns that provide insight. It allows us to communicate complex relationships that we otherwise would not see. From telling the story of drone strikes in Pakistan, to the drought problem in California, to 311 call patterns in New York City, we’ll look at the process of visualizing data and using it to tell stories and improve understanding.
1:00 PM
Members' Working Lunch — Georgian Room
2:15 PM
Mobile Data: The Pulse of the Networked Society
Gordon Castle, Vice President, Head of Industry Area Mediacom, Ericsson
Screen size matters: tablet users spend 50 percent more time watching videos online than average mobile broadband users. Sports are already a key driver of video use, and real-time viewing, sharing, and social networking surrounding sports have become integral parts of sports events. The combination of smartphones, apps and mobile broadband coverage forms an enhanced spectator experience. Smartphone subscriptions are set to more than double by 2020; 70 percent of the world´s population will have a smartphone by then. Overall, there will be 26 billion connected devices by 2020, but smartphones will account for 80 percent of mobile data traffic. These are just a few of the surprising results in the annual Ericsson Mobility Report.
2:40 PM
Data: A Love Story
Christian Rudder, Co-founder, OKCupid
For centuries, we’ve relied on polling or small-scale lab experiments to study human behavior. Today, a new approach is possible. As we live more of our lives online, researchers can observe us directly, in vast numbers, and without filters. Data scientists have become the new demographers, and those with access to proprietary data are in a position to discover truly new facts about ourselves. The dating site OKCupid is sitting on top of one such database, with information that not only can monitor the gap between people’s stated preferences and the choices they actually make, but also such things as human migration over time—how groups of people move from certain small towns to the same big cities across the globe.
3:20 PM
Conference Reflections
Bob Lucky, TTI/Vanguard Advisory Board
4:00 PM
Close of Conference

home about us activities and deliverables contact faqs copyright