Now Live

Schedule

Opening Remarks08:20 - 08:30The Ballroom
Reinforcement Learning In The Wild & Lessons Learned08:30 - 09:00The Ballroom
Reinforcement Learning (RL) is an area of AI to optimize sequential actions based on the perceived state of the environment and the feedback from applying such actions in order to maximize a long term objective. This generic definition extends to many interesting application areas. For instance, the environment can be a user behavior (marketing, personalization, mobile health), robots (industrial robots, autonomous vehicles), market of actors (financial market, Adwords bidding, etc), games (Atari games, Go, etc..), conversation and AI assistants, IoT, biology, or agriculture. We have seen applications of RL in all those areas with varying degrees of success. I will be talking about such applications, mainly in the marketing and manufacturing verticals, and construct a common design framework for the use cases from a product management point of view. I will share some lessons learned from applying RL in the Wild versus applying RL in the Lab and talk about points such as time variant environment and coherence time, objective & subjective reward design and pitfalls, and the high cost of exploration.
Using AI to provide insights and recommendations from activity data09:00 - 09:30The Ballroom
In the customer age, being able to extract relevant communications information in real-time and cross reference it with context is key. This presentation will explain how Salesforce is using data science and engineering to monitor Salespeople activities in real-time and surface insights and recommendations across products. Salesforce is developing Einstein which is an artificial intelligence (AI) capability built into the core of the Salesforce Platform. Einstein helps power the world’s smartest CRM to deliver advanced AI capabilities to sales, services, and marketing teams – helping them discover new insights, predict likely outcomes to power smarter decision making, recommend next steps, and automate workflows so users can focus on building meaningful relationships with every customer. In this presentation, Alexis will explain how Salesforce Einstein Activity Platform combines activity data (such as emails) with contextual knowledge from users and CRM data to provide real time insights and recommended actions. Alexis will go over use cases, high level architecture and how a variety of technologies (data engineering, data science, graph processing, NLP, machine learning and deep learning) are combined together to support Salesforce applications.
Geometric Deep Learning For Movie Recommendations09:30 - 10:00The Ballroom
Movie studios face a complex landscape. AI won’t make movies better any time soon, but can help make them more successful. Although AI research for digital content has exploded, there is a big research gap for offline transactions in general, and theatrical movies in particular (e.g., which AI models to use, how to use the models, etc). In our recent work we explore deep models to help organize content & market data, and surface it in a language that is useful for content production. Downstream, we explore how deep models can ‘plug’ into tech & media platforms to create adaptive campaigns to drive box office outcomes.
Moneyball In Hollywood10:00 - 10:30The Ballroom
British Prime Minister Benjamin Disraeli once said, “There are three kinds of lies: lies, damned lies, and statistics.” Hollywood is slowly coming around to data-driven decision-making, but skepticism towards quantitative findings still lingers… This presentation will discuss data usage and data silos at studios. Additionally, AI/ML entertainment examples will be shared to show how an improving analytics culture is providing actionable insights to mitigate risk and drive profits.
Roundtable Surgery10:30 - 10:45The Ballroom
In this session, the 'patient' will share he/her biggest pain experiencing at work. 'Table surgeons', aka the rest of the table, will diagnose the problem and treat the issue.
Coffee Break10:45 - 11:15The Foyer
Natural Language Processing - Concepts, and Comparison of off-the-shelf Platforms11:15 - 11:45The Ballroom
This talk introduces concepts of Natural Language Processing (NLP), including discussion on Intents, Utterances, Objects, Slots, and Contexts. This talk will also discuss prominent AI methods used in NLP, and various parameters used to benchmark quality and performance of NLP methods. This discussion will cover some challenges in NLP implementations. Finally, this talk will take a practical approach to comparing various public NLP platforms including Api.ai (DialogFlow), Microsoft LUIS, IBM Watson Conversation Service, and Facebook's Wit.ai. The discussion will cover pros and cons of these four platforms, and will provide some example applications.
Panel Session: Challenges & Solutions to Innovative Use of AI in Business11:45 - 12:30The Ballroom
Romit Jadhwani, Head of Ads Analytics & Business Intelligence, Pinterest; Tish Shute, Director, AR/VR, Corporate Strategy & Technology Planning, Huawei; Amit Srivastava, Director, Research Engineering, ebay; Eru Akyuz, AI & Advanced Analytics - Cloud Solutions Architect, Microsoft
The Rapid Evolution and Future of Machine Perception12:30 - 01:00The Ballroom
With the advent of deep learning, our ability to build systems that derive insights from perceptual data has increased dramatically. Perceptual data dwarfs almost all other data sources in both its richness and its sheer size. This poses some unique challenges that have forced learning systems to evolve. This technical progress has enabled learning systems to be adopted in mainstream consumer products across the industry, such as Google Photos and YouTube, where learning systems have clearly proven their usefulness. In this talk, we’ll review the key ingredients of recent progress in machine perception. We’ll also explore the substantial gaps that still need to be filled, and highlight some emerging applications that illustrate the potential future impact of this technology.
Lunch01:00 - 02:00The Foyer
How Machine Learning Is Changing The Face Of Cyber-defense02:00 - 02:30The Ballroom
Cyber-defense is undoubtedly the fastest changing arena for Internet security; with that rapid change being driven almost exclusively through advances in machine intelligence and the widespread adoption of machine learning. What used to take days of work sifting through logs and correlating alerts per incident, is now collated and classified in real-time, and is on the cusp of permanently replacing a sizable chunk of the information security workforce. The security pendulum has swung in the defenders favor, but the successes of machine intelligence have not got unnoticed by the attackers. How long before attack, defense, and remediation, are exclusively the remit of the machines?
The Future Of AI02:30 - 03:00The Ballroom
There are a number of perspectives and predictions on the future of AI. This talk will cover some key view points and explore the past, present and future of AI. It will explore some of the key advances in AI and AI algorithms and make educated guesses about how the field of AI will evolve over the near, medium and long term. This is a popular science talk that is meant to both entertaining and informational.
Roundtable Surgery03:00 - 03:15The Ballroom
In this session, the 'patient' will share he/her biggest pain experiencing at work. 'Table surgeons', aka the rest of the table, will diagnose the problem and treat the issue.
Coffee Break03:15 - 03:45The Foyer
Panel Session: How Can Your Business Use AI & What're The Best Practices03:45 - 04:30The Ballroom
Azita Arvani, Head of Innovation Partner & Venture Management, Nokia; Sohini Roy Chowdhury, Sr Specialist, Deep Learning Engineering, Volvo Cars; Srinivasan Seshadri, VP, AI Led NLP & Search, Target
WHEN AI GETS CREATIVE, HOW DO COMPANIES STAY ON TOP04:30 - 05:00The Ballroom
ONCE RELEGATED TO ROTE TASKS, ARTIFICIAL INTELLIGENCE IS SLOWLY MOVING INTO CREATIVE FIELDS, WHICH HAVE TRADITIONALLY BEEN THE DOMAIN OF HUMANS… FROM THE GOOGLE ARTIST IN RESIDENCE WHO BUILDS ART-GENERATING SOFTWARE WITH MACHINE LEARNING TO AN ALGORITHM THAT MIMICS THE WORKFLOW OF A LANDSCAPE PHOTOGRAPHER, PRODUCING PHOTOS THAT ARE NEARLY AS GOOD AS PROFESSIONAL SHOTS.
Adaptive Experiences using AI in the Digital Home05:00 - 05:30The Ballroom
Closing Remarks05:30 - 05:35The Ballroom
Networking Drinks05:35 - 07:00The Foyer

Tickets expired

Final Release Ticket

$445