The explosion of machine learning and its many applications has motivated a variety of new
domain-specific architectures to accelerate these deep learning workloads. The Groq Tensor Streaming Processor (TSP) is based on a deterministic instruction set architecture (ISA) with a single large core. The ISA exposes temporal information indicating the number of cycles the instruction requires to produce the output stream — its functional latency. Determinism allows the compiler to reason about program correctness and track the exact spatial and temporal position of every tensor on the chip. Events in a deterministic system cannot be permuted by the underlying hardware — that is, the total program order is the interleaving of individual instruction queues of each functional unit. This total ordering is entirely software controlled and the underlying hardware cannot reorder these events and they must complete in a fixed amount of time. This has implications for hardware design: it removes use of hardware interlocks to coordinate between functional units, or use any “reactive components” in the data path such as arbiters, caching agents, replay or retransmission mechanisms, etc. It also has several consequences for system design: zero variance latency, low latency and high throughput at batch size 1, reduced total cost of ownership (TCO) for data centers with diverse service level agreements (SLAs) and the ability to scale to large training and inference systems without traversing networking switches. In this talk we discuss the TSP and the design implications of its architecture.
Chief Security Officer of FortressSecure
Federico is Chief Security Officer of FortressSecure, a company disrupting multi-cloud data security through AI assisted vault-less encryption innovation. He is a serial entrepreneur with several successful exits in his past, including Internet Graffiti: web development company, sold to the Cabassi (Bastogi/Brioschi) Group, one of the largest Italian corporate groups. K2SXS: security software development company, acquired by merger into 9Proof Italia, later turned into Extenua. He is a ethical hacker (fka DDT) and private IT security investigator; performed security assessment of various top corporations, like Siemens and the Italian National Highway Authority (Società Autostrade). He also served as a professor of operating systems security at the University of Milan and assisted both Italian government agencies and enterprises on data and internet security issues. One of the solutions he developed for DIGOS (the Italian national anti-terrorism police) is to visualize the criminals’ activity without being seen/identified; this effort has led to identifying and capturing pedophiles.
Software Engineer (Computer Vision), DeepNorth
Priyanka Makhijani received B.S. in Computer Science from Manipal University Jaipur, India in 2016 and M.S. in Computer Science from Illinois Institute of Technology, Chicago in 2018. She has experience working in Tunis, Tunisia, Bangalore, India during her Bachelors’s. She joined the R&D department at Siemens Healthineers working as a Deep Learning Engineer followed by Research Assistantship at Linguistic Cognition Lab at Illinois Institute of Technology working on ML techniques on XAFS data. Currently, she is working as Software Engineer focused on Computer Vision problems at DeepNorth Inc. her work is mainly related to image analysis, object detection and segmentation problems in research space in applicative domains.
Facial Recognition Systems
Facial recognition as a computer application is an emerging trend leveraging current computer vision applications with AI-based technology. This emerging technique has reshaped the research landscape of face recognition since 2014, launched by the breakthroughs of Deepface and DeepID methods. Since then, deep face recognition (FR) technique, which leverages the hierarchical architecture to learn discriminative face representation, has dramatically improved the state-of-the-art performance and fostered numerous successful real-world applications. It is extensively used in facial bio metrics, engaging customers in a store by reading facial expressions and emotions, video surveillance, targeted advertising, tagging and searching pictures online etc. However, there are potential deficiencies of the current methods and several future directions which needs attention for further research and improvement.
Machine Learning Engineer, Weights
Lavanya is a machine learning engineer at Weights and Biases which builds performance and visualization tools for machine learning teams and practitioners. She began working on AI 10 years ago when she founded ACM SIGAI at Purdue University as a sophomore. In a past life, she taught herself to code at age 10, and founded the machine learning startup Dataland. She’s driven by a deep desire to understand the universe around us better by using machine learning. You can find her on twitter @lavanyaai.
Neural networks are powerful beasts that give you a lot of levers to tweak! The sheer size of customizations that they offer can be overwhelming to even seasoned practitioners. In this talk I’ll give you a framework for making smart decisions about your neural network architecture! We’ll explore lots of different facets of neural networks in this talk, including how to setup a basic neural network (including choosing the number of hidden layers, hidden neurons, batch sizes etc.) We’ll learn about the role momentum and learning rates play in influencing model performance. And finally we’ll explore the problem of vanishing gradients and how to tackle it using non-saturating activation functions, BatchNorm, better weight initialization techniques and early stopping.
– Based On: https://www.wandb.com/articles/fundamentals-of-neural-networks
– Code: http://bit.ly/keras-neural-nets
VP of Engineering, Groq
Michelle Tomasko is leading the hardware and software engineering effort at Groq. She has twenty years of technical leadership in the silicon industry and a proven track record for propelling engineering teams to success on ambitious programs with aggressive schedules. She has a passion for data parallel computation, microprocessor architecture and low power designs.
In the past, she worked for leading technology companies including NVIDIA, Google Consumer HW, and Transmeta. At NVIDIA, she drove all aspects of software for the leading GPU architectures. Michelle also led the SW effort for NVIDIA’s first consumer Android device, Shield. She delivered Google’s first Machine Learning/Image Processing Accelerator System on a Chip, the Pixel IPU for the Pixel 2 phone.
Director, Venture Capital Venture Capital at Micron, Inc.
Gayathri Radhakrishnan is currently part of the investment team at Micron Ventures, investing from $100M AI fund. She invests in startups that are leveraging AI/ML to solve critical problems in the areas of Manufacturing, Healthcare, Automotive and AgTech. Prior to that, she brings 20 years of multi-disciplinary experience across product management, product marketing, corporate strategy, M&A and venture investments in large Fortune 500 companies such as Dell and Corning and in startups. She has also worked as an early stage investor at Earlybird Venture Capital, a premier European venture capital fund based in Germany. She has a Masters in EE from The Ohio State University and MBA from INSEAD in France. She is also a Kauffman Fellow.
Prior to her own startup, Uma led chip design projects in various application spaces from microprocessors to high speed fiber optic communications, through startups and in-house incubators to large public companies. Her 18+ years of hands on experience has rare combination of not only application spaces, but of technology use and design domains. She is also an international artist and instructor who travels worldwide to teach art and to talk about art-engineering balance. She serves on the Executive Board of a global 100K+ member non-profit that promotes sketching on location. An artist-engineer combo from Bay Area, she believes that for elegant solutions, less is more and that it holds true in both engineering and art. Uma holds a M.S. Electrical Engineering from Stanford University.
Founder, AI Technology & Systems
Bio
Rohit Sharma is an engineer, author and entrepreneur. He has published many papers in international conferences and journals. He has contributed to electronic design automation domain for over 20 years learning, improvising and designing solutions. He is passionate about many technical topics including machine learning, deep learning, VLSI characterization, analysis, and modeling. It led him to design and architect several design automation products. He wrote a book on Machine Intelligence in Design Automation in 2018. He has been disseminating information on the use of machine intelligence in design automation since 2017 with his book, blogs, opens source code, contests, webinar, symposium, and other collateral. He currently works for AI Technology and Systems (www.ai-techsystems.com).
Abstract
Dnn Compiler is targeted towards devices with small form factor like microcontrollers, which are part of all sorts of household devices: think appliances, cars, and toys. In fact, there are around 30 billion microcontroller-powered devices produced each year. They’re cheap, require very little energy, and are very reliable. By bringing deep learning models to tiny microcontrollers, we can boost the intelligence of billions of devices that we use in our lives, without relying on expensive hardware or reliable internet connections. Imagine smart appliances that can adapt to your daily routine, intelligent industrial sensors that understand the difference between problems and normal operation, and magical toys that can help kids learn in fun and delightful ways.
Rambus Senior Director, Product Management
Ben Levine has over twenty years of experience in security, ASIC design, and system architecture. Dr. Levine joined Cryptography Research, Inc, in 2008, prior to its acquisition by Rambus in 2011. His work there has included content protection, anti-counterfeiting, anti-tamper, and hardware roots of trust. Prior to joining Cryptography Research, he was Director of Engineering at Rapport, Inc., a fabless semiconductor company commercializing massively parallel reconfigurable computing architectures that he helped develop at Carnegie Mellon University (CMU). Before his work in industry, he was a research faculty member at CMU and an Assistant Professor at the University of Pittsburgh. He has a B.S and M.S in Electrical Engineering from the University of Tennessee, Knoxville, and a Ph.D. in Electrical and Computer Engineering from CMU.
Dedicated accelerator hardware for artificial intelligence and machine learning algorithms are increasingly prevalent in data centers and endpoint devices. These accelerators handle important data which has value and must be protected. Many security threats exist that can compromise these assets. Fortunately, there are security techniques which can mitigate these threats. This talk will outline the threats and mitigations and describe the security building blocks needed. It will then show ways to practically implement these building blocks in AI accelerators in endpoint devices or in the cloud.
Oracle Cloud Solution Engineer
Girish Adiga is a Consulting member technical staff at Oracle, working on Autonomous Health features in Oracle Autonomous Database product. Oracle Autonomous Health Framework uses applied machine learning techniques to proactively identify issues before they affect the health of database clusters and take actions to fix or alert the administrators. He has been working for database product development team at Oracle for past 18 years, currently focused on anomaly detection and predictive analytics to enable quick root cause analysis for database issues using machine learning.
Senior Director of Transformational Technologies, Oracle
Anant Kadiyala is Senior Director of Transformational Technologies at Oracle, where he leads business innovation and enterprise solutions with ML/AI, IoT, blockchain, and VR/AR. Anant is a specialist in enterprise digital transformation and applied innovation. He has published numerous articles and delivered keynotes at conferences on applying emerging technologies for business and social good. Earlier, Anant built a consulting practice that specialized in analytics, IoT, PaaS, and middleware solutions for the Fortune 500. His team won multiple industry awards for building visionary product platforms for customers. Anant is based out of San Francisco Bay Area. Twitter: @akadiyala
Technologies like AI/ML, IoT, blockchain, and cloud are compelling and powerful in their own right. However, when used together they slash the ‘transactional costs’ and cause business ecosystems to quickly reorganize. With a focus on AI/ML, this session explores the new digital infrastructure underpinnings that are rapidly emerging, and unpacks a few new business affordances they enable. Are we ready for the future?!
Tractica Principal Analyst
Professional with expertise in computer vision, AI, video analytics, 3D and related technologies. Excellent at converting technical lingo into business implications, marketing, business development, business strategy. Blogger. Frequent speaker in conferences. Start-up advisor.
– Managed business size ranging up to $200M.
– Track record of growing business units (Adobe Flash Scaling Partner) and start-up (Poseidon –
multimedia acceleration).
– Closed several multi-million dollar business deals.
– Product and solutions marketing and business development background. Venture capital and market research experience.
During this session, Mr. Joshi will share insights on market trends and opportunities for creating AI/ML Accelerators.
Technology evangelist, Intel
Sujata Tibrewala is an Intel community development manager and technology evangelist who defines programs to enable ecosystem developers to drive AI/ML. With OneAPI being the latest program she has taken the responsibility of evangelizing, she is also a co-chair for IEEE Edge Automation Platform Roadmap, for Beyond 5G Technology Roadmap. Under her leadership Intel Network Developer Evangelism program was nominated for Network Transformation Awards 2018, and received Edison award and Network Developer Dynamo award at Intel. She is a frequent presenter at various IEEE and industry conferences in SDN/NFV, Director at Silicon Valley Engineering Council and TSC chair for Documentation Sub-committee Akraino. Sujata has worked at several companies, including CISCO, Agere, Ericsson, Avaya, Brocade, leading all phases of diverse software technology projects such as an SDN open flow implementation, TCP/IP/Ethernet/VLAN forwarding software development on CISCO switches, and network processors and cloud deployments using virtualization technologies. She has a Masters from IISc Bangalore and Bachelors from IIT Kharagpur and has completed an Executive Women Leadership Program from Stanford.
5G, Edge, cheap computing means deluge of micro Data Centers spawning close to the user locations. This open up a whole new model of delivering apps and services across multiple architectures and implementations in the industry. This infrastructure also needs to be scalable enough to distribute functions to serve billions of parallel devices and tasks at the edge. . We talk about how workloads and applications will drive the edge architecture for Domain Specific Applications (DSA) where by various technology enablers such as AI, ML and oneAPI- a set of developer tools that provide a unified programming model that simplifies development for workloads across diverse architectures, can create a level playing field for the developers of these applications by letting them take full advantage of multi architecture platforms easily. Various DSA’s include Green Applications aka Rideshare, Green champion social app, Real time location-based services, Social networking applications, Real Time XR based use case. With the push towards 5G and the promise of connecting the unconnected technologies such as ad Hoc Network based apps where any node can opt in to connect other nearby nodes have huge potential to bridge the Digital Divide. We will conclude by talking about various Open source efforts in the edge space in general and Akraino Edge Stack in particular which aims at making the edge a more accessible space for software developers on the edge.
See also the Panel in which Sujata is a panelist.
CEO, Greenlight Insights
Clifton is CEO and Founder of Greenlight Insights, a global leader in market intelligence about the future of display technologies. Clifton is responsible for all research and client services at Greenlight Insights. Prior to founding Greenlight Insights, Clifton was growth and monetization analyst at Snap, the leading image messaging and multimedia platform. Previously, Clifton was Head of Business Operations at AddLive, the web-based real-time communication platform acquired by Snapchat in 2014. Clifton has also worked as a Business Analyst for Bell Canada, a Canadian telecommunications and media company. Clifton has a B.A. in Economics from Harvard University and an M.B.A. from Harvard Business School.
During this session, Mr. Dawson will share insights on market trends and opportunities for creating next-generation entertainment experiences based on recent advances in computer vision technology.
Infosys Ltd Principal Business Consultant
As a product leader with a technical background, Soham brings a record of success and expertise with a range of companies, from startups to the Fortune 500. For more than 11 years, he built and managed software and hardware products with excellent user experience as his top priority. He has contributed to building, launching, and executing successful business models, securing large partnership deals, crafting product strategy, building teams, and executing go-to-market. Beyond expertise, he successfully combines a world-class technical background with business acumen and strong communication skills a key to consistently building great products, closing significant deals, and securing key partnerships. More recently he was finalist at Falling Walls contest at Stanford University on Blockchain, AI technologies.
The recent, exponential rise in the adoption of Internet of Things (IoT) devices and Blockchain- AI technologies has also reached Agriculture and Food (Agri-Food) supply chains, drumming up substantial research and innovation interest towards developing reliable, auditable and transparent traceability systems. Current IoT-based traceability and provenance systems for Agri-Food supply chains are built on top of centralized infrastructures, and this leaves room for unsolved issues and significant concerns, including data integrity, tampering and single points of failure. Blockchains, the distributed ledger technology underpinning cryptocurrencies such as Bitcoin, represent a new and innovative technological approach to realizing decentralized, trustless systems. Indeed, the inherent properties of this digital technology provide fault-tolerance, immutability, transparency, and full traceability of the stored transaction records, as well as coherent digital representations of physical assets and autonomous transaction executions. This talk presents AgriBlockIoT, a fully decentralized, blockchain- based traceability solution for Agri-Food supply chain management, able to seamless integrate IoT devices producing and consuming digital data along the chain. To adequately assess AgriBlockIoT, first, we defined a classical use-case within the given vertical domain, namely from-farm-to-fork. Thirty percent of the global food supply is wasted – the retail equivalent of $1 trillion of food each year, says the Food and Agriculture Organization(FAO) of the UN. In addition to the retail cost of food lost, another $700 billion is also thrown out in natural resources, including $172 billion in wasted water, $42 billion in cleared forest and $429 billion in related greenhouse gas costs. How IoT can generate real-time data, Enterprise Blockchain can store IoT based generated real-time data & AI will help for future prediction & analytics. How IoT-Blockchain-AI can together be used in a more recycle industry to produce natural composite, animal feed & save so many resources. All this is possible due to the synchronous integration of the above technologies together with efficient hardware node together with sensors in the edge. Finally use cases covering above will be presented as proof of concept.
Group Product Manager, Service Now
Product management leader with 16 years of experience in the tech industry with startups, mid-size and high profile public companies. Manjeet is a strategic thinker and deeply passionate about learning customer problems with full empathy and providing elegant solutions. Manjeet has a great track record of delivering high impact enterprise and consumer-facing products from ideation to launch to growth. He holds a Bachelor degree in Computer Engineering and an MBA in Product Management & Marketing from Santa Clara University.
Cloud applications, artificial intelligence, machine learning, data insights, rapid prototyping, design thinking, and faster decision making are becoming more and more significant in the daily life of a manager. Basic physical, digital, and biological technologies are intersecting to create large scale system changes, altering the very fabric of our social system. More than ever, the manager is faced with the challenges of proactively designing the systems of the future.
Executive Analyst, Rainq Strategies
Ram Appalaraju has held senior executive roles (CEO, CMO, SVP) in product management and marketing large enterprises (GE, HP, Cisco) as well as start-ups (8kpc, Meru Networks, Isilon) in over two decades. During his tenure he has successfully developed complex product strategies and business models to generate profitable growth in the companies he worked for. He is experienced in several core areas of new technologies such as cloud computing, distributed application services, IT orchestration, AIOPS, etc. He is currently at Rainq Strategies as an Analyst assisting clients develop and adopt AI based product strategies with deep considerations for innovation, technology and business outcomes.
It is a question of time that AI-based products and services will be a pervasive aspect of consumer and business services by 2025. Several established companies and start-ups have embarked on product/services development, capitalizing on plethora of tolls and platforms that are available and continue to grow. While innovating a key technically challenging solution is a daunting and difficult task, it’ll be incomplete without innovative business aspects that are relevant for the present and the future. Bringing to market AI based products and services will require a fresh thinking. Product management is becoming a strategic function and traditional ways of developing MRD’s and PR’s are insufficient to complete a true business-driven product conception. The lecture will outline key frameworks for product managers and entrepreneurs to consider in developing such a wholesome product strategy.
See also the Panel in which Ram is a panelist.
Lead Machine Learning & Big data Engineer Verizon Smart City
Overall 18+ yrs in Product development for Enterprises and Startups. Last 8 yrs working in Data Science, ML and Big Data Engineering for eCommerce, CRM, IoT, Networking, Smart City and HealthCare use cases.
Deep Learning Algorithms are applied in eCommerce domain for solving many different problems mining product review , understanding customer search intent, finding best advertisement to display, optimizing item delivery , boosting page rank and many more. eCommerce is driven by insights gained from product items and users. Some sample use cases are Analysis of Search Query, Product Delivery, Review Mining, Advertise placement, Catalog categorization, Image analysis, Search optimization
Deep Learning Engineer, Nvidia
Subhashree Radhakrishnan is a Deep Learning Software Engineer in the Intelligent Video Analytics Team (Metropolis AI) at NVIDIA. She earned her Masters Degree from Virginia Tech with specialization in AI and Computer vision. At NVIDIA, she works on developing models for Smart City applications on edge devices that will make our surroundings smart and safe. Her research focus was on visual understanding in videos and intersection of image and language domain during her Masters. She is particularly interested in leveraging AI for social good and is a strong advocate of Women in Tech. She is currently a mentor for Generator program at MIT CSAIL for undergraduate students working on projects with societal impact.
Chief Technologist, Semiconductor BU, ANSYS Inc.
Norman Chang co-founded Apache Design Solutions in February 2001 and currently serve as Chief Technologist at Semiconductor BU, ANSYS, Inc. He is also currently leading the effort of applying ML/DL at ANSYS. Dr. Chang received his Ph.D. in Electrical Engineering and Computer Sciences from University of California, Berkeley. He holds fourteen patents and has authored over 50 technical papers. He is currently in the committee for ESDA-EDA and IEEE P2401.
In this talk we will discuss how AI and machine learning can improve simulation, and how simulation can improve AI and ML.
Senior Software Engineer, Nvidia
Madhumitha is a senior deep learning software engineer in the TensorRT team at NVIDIA. She earned her Masters degree in computer engineering with specialization in computer systems and computer architecture at Carnegie Mellon University. Prior to that, she was a software developer at Goldman Sachs technology. At NVIDIA, she works on optimizing and deploying diverse deep learning models for inference on NVIDIA GPUs. She is interested in exploring the world at the intersection of software development, deep learning and computer architecture. She is passionate about empowering women in technology and believes that women have a key role to play in shaping the future of AI.
Deep learning inference optimization and deployment using TensorRT
NVIDIA TensorRT is a platform for high-performance deep learning inference. It includes a deep learning inference optimizer and runtime that delivers low latency and high-throughput for deep learning inference applications. TensorRT-based applications perform up to 40x faster than CPU-only platforms during inference. With TensorRT, you can optimize neural network models trained in all major frameworks, calibrate for lower precision with high accuracy, and finally deploy to hyperscale data centers, embedded, or automotive product platforms. In this session, I will talk about why inference is a huge market opportunity, why optimizing inference is important for AI to be effective and what are some common techniques used to optimize deep learning inference.
CTO Symantec Corp
Ashok has led products from startups to Fortune 150 companies, both consumer and enterprise focussed. He has led products across CyberSecurity, Logistics, ECommerce, User Growth, Social Networks, Machine Learning, Payment Platforms, Big Data and Analytics and large scale systems. He enjoys leading amazing products which years later I can look back and still feel proud of.
Ashok’s technical interests are Machine Learning, Fractal Theory, NLP, CyberSecurity (Threat and Information Protection), Machine Learning, NLP, Graph Theory, Operations Research, Optimization Theory, Approximation Algorithms, Game Theory, in Industrial Internet (IIOT) and IOT, Complex Event Processing Systems. Unrelated recent interest is fake news detection.
Ashok continues to invent new technology and my 22 US patents (9 pending) cover a wide variety of domains.
Cybersecurity on Endpoints for long has been about killing when we are highly confident that something is bad. If the false positive has to be dramatically low, by definition the product has to let-go what is quite suspicious but yet not suspicious enough to cross the very low false positive threshold. This is a self inflicted trap for Cyberdefense. This session is about the forgotten dimension – machine learning to learn the user, hisher activity sequences, what they never do and personalize security. The longer the user stays with the product, the more secure the user becomes. It literally learns the user much like a spouse would from day-to-day from hours of activity with a computer. This approach has significant value – in switching the playground on the attacker. Every endpoint is different, which is in stark contrast to today’s world where majority of endpoints run SEP in its defaults or Windows Defender in its default settings. This frequent defaulting makes a very strong static security against a dynamic adversary. Regardless of how good the static (default security) maybe the attacker can test infinite number of times, across machines in the cloud until they know some mutations will succeed. Current attacker test beds include subscription services to perform live updates constantly, without submitting samples to the cloud and then performing mutations.
Given that increasingly more systems are reliant on machine learning we expect a sharp rise in standard adversarial techniques within the test bed, provided as a service. Expected adverserial libraries will likely use with append attacks (recent complete bypass of a cybersecurity vendor), prepend attacks, substitution, permutation, insertion attacks on the machine learning models of the common antimalware. This session is about personalized security where every endpoint is different because every person is different then we are attacking the very testing workflow of the attacker. Making every attack prohibitively expensive, killing the portability of the attacks themselves. This I expect would dampen retail attacks significantly. This is a attack on economics of the attacker and the testbeds for attackers. Default is the enemy of the Cyberdefense, and this model has no defaults at all. In conclusion personalized security is a overlay of small-data-learning (on data of the one) on top of the global big-data-learning.
***
See also the Panel in which Ashok is a panelist.
Distinguished Engineer – Machine Learning & Computer Vision
Dr. Gunnam is an innovative technology leader with vision and passion who effectively connects with individuals and groups. Dr. Gunnam’s breakthrough contributions are in the areas of advanced error correction systems, storage class memory systems and computer vision-based localization & navigation systems. He has helped drive organizations to become industry leaders through ground-breaking technologies. Dr. Gunnam has 75 issued patents and 100+ patent applications/invention disclosures on algorithms, architectures and real-time low-cost implementations for computing, storage and computer vision systems. He is the lead inventor/sole inventor for 90% of them. Dr. Gunnam’s patented work has been already incorporated in more than 2 billion data storage and WiFi chips and is set to continue to be incorporated in more than 500 million chips per year.
Dr. Gunnam is also a key contributor to the precise localization and navigation technology commercialized for autonomous aerial refueling and space docking applications. His recent patent pending inventions on low-complexity simultaneous localization and mapping (SLAM) and 3D convolutional neural network (CNN) for object detection, tracking and classification are being commercialized for LiDAR + camera-based perception for autonomous driving and robotic systems.
Dr. Gunnam received his MSEE and PhD in Computer Engineering from Texas A&M University, College Station. He is world-renowned for balance between strong analytical ability and pragmatic insight into implementation of advanced technology. He served as IEEE Distinguished Speaker and Plenary Speaker for 25+ events and international conferences and more than 3000 attendees in USA, Canada and Asia benefited from his lecture talks. He also teaches graduate level course focused on machine learning systems at Santa Clara University.
VP ML & AI and Distinguished Engineer at MINDBODY, Inc.
Prasad Saripalli serves as the Vice President of ML & AI and Distinguished Engineer at MindBody Inc – a portfolio company of Vista which manages the world’s fourth-largest enterprise software company after Microsoft, Oracle, and SAP. Earlier, he served as VP Data Science at Edifecs, an industry premier healthcare information technology partnership platform and software provider, where we built Smart Decisions ML & AI Platform with ML Apps Front. Prior to joining Edifecs, he was chief technology officer and VP of engineering at Secrata.com, which provides military-grade cloud security solutions. Previously, he worked as chief technology officer and executive VP at ClipCard and as chief architect for IBM’s SmartCloud enterprise. He also served as principal group program manager on Microsoft’s client virtualization team, which was responsible for shipping Virtual PC on Windows 7, and as a dev manager on the Citrix group that built Citrix Presentation Server (now Citrix XenApp).
Prasad has a master’s in computer science from Washington State, doctoral training in engineering and computer science from the University of Florida and post-doctoral training from the University of Texas.
AI-enabled Marketplaces for Wellness, Beauty and Fitness: Discoverability
Retail e-commerce sales worldwide are projected to double in the next 3 years to $ 6.54 trillion. Convenience of always-on service availability, price comparison, better sales and discounts, greater choice enabled by worldwide, multi-cultural varieties of goods and services and delivery to homes free of charge are fueling a steady growth of e-retail. However, discoverability of online products and services continues to be a significant cognitive problem, known as the paradox of choice. Users abandon an online store visit for competitors after 2 to 3 seconds of unsatisfactory discovery experience. Success of an online marketplace lies in how quickly users find the product they need: quick, efficient discovery of the right product is shown to lead to higher conversion rate, better customer satisfaction, engagement and retention. Poor product discovery often leads to additional economic and environmental costs due to cancellation of orders. Several innovations such as visual search are emerging to address the need for better discovery. Gartner estimates that by 2020 30% of all searches will be conducted ‘query-less’. Recommender systems, which seek to present the perfect product to the users based on detailed profiles of users’ shopping behavior are unable to meet the need fully, and online storefronts like Amazon and Alibaba have established physical retail presences to augment the online stores and provide a more satisfactory buying experience. Discoverability in digital marketplaces is not entirely an information search problem; it is multi-faceted and is strongly influenced by the inventory, tagging and on-boarding of content, the presentation layer, search engine design and optimization, and the users’ domain knowledge and cognitive abilities as well.
If search and discoverability of basic, essential goods and services such as clothing poses such significant challenges to both the consumer and the retailer, it is understandable that the problem of discoverability is even more daunting in the context of complex and highly specialized, life critical services such as Healthcare, Fitness and Wellness. Wellness – the active pursuit of activities that promote physical and mental well-being including fitness, beauty and integrative health, is a $4.5 Trillion global economy of services. In this session, we will first survey how the modern trends of service-oriented and ubiquitous computing are disrupting the Wellness industries, followed by an in-depth treatment of the key ideas and challenges, algorithms and applications of product/service search and discoverability enabled by AI and ML in modern digital service platforms, using Wellness as a case study.
VP and Chief Technical Officer for UL’s Connected Technologies businesses
Tom Blewitt is VP and Chief Technical Officer for UL’s Connected Technologies businesses. He is based in New York, USA and has spent the majority of his 40+ year career in engineering and engineering management roles involving product safety.
A William Henry Merrill Society member and Corporate Fellow, Mr. Blewitt is responsible for technical consistency, integrity and engineering quality for application of UL’s standards and for certification activities. He has extensive experience in the development of US, regional and international safety standards with recent focus on autonomy, robotics and security of devices connected to public networks. He was recognized with the American National Standards Institute (ANSI) Meritorious Service Award and the International Electrotechnical Commission (IEC) 1906 Award. He has been a member of the IEC Advisory Committee on Safety (ACOS).
Mr. Blewitt has published numerous articles and has regularly presented to trade association and government audiences on product safety, standardization and international safety standards harmonization. He holds a BSEE and MS Mgt. from Polytechnic Institute of NY. He is also a licensed Professional Engineer in the State of New York and a member of the U.S. National Fire Protection Association.
Understanding the need for standards in testing for autonomous system safety. UL 4600 and how it applies to cars, drones, and other vehicles. Who is involved in UL 4600 and what is the process? The technical background on how testing/verification works.
