Intel on AI

In this Intel on AI podcast episode: Recording live at the Intel AI Summit in San Francisco California, Deepak Dinkar Senior Practice Manager at Wipro Technologies, joins the Intel on AI Podcast to discuss how Wipro is working with Intel AI technologies to drive a wide array of innovative solutions. Deepak discusses how the Wipro Pipe Sleuth solution uses artificial intelligence (AI) to automatically process video scans of municipality sewer and water pipes to identify and mitigate pipe leakage, breakage, and blockage, which could result in property damage or safety hazards. He mentions how DC Water in Washington DC is already seeing benefits in the reduction of time it takes to inspect and maintain their piping infrastructure by using Pipe Sleuth. Deepak also highlights other innovations from Wipro including their surface crack detection solution which uses AI to identify potential defects in concrete structures enabling inspectors to more rapidly address safety concerns. He also outlines the Wipro medical imaging solution which utilizes the Intel Distribution of OpenVINO toolkit to speed up analysis of medical images helping to diagnose diseases from CT and x-ray scans. Lastly, Deepak discusses how being a member of the Intel AI Builders program has helped Wipro address different customers and verticals to create new and innovative solutions. To learn more, visit: wipro.com Visit Intel AI Builders at: builders.intel.com/ai

2 Joulu 20199min

In this Intel on AI podcast episode: Enterprises today have exploding volumes of data which is growing the scale and complexity of data centers and often result in unplanned IT downtime. This disrupts mission-critical operations, causes loss of data, and impairs application services. Atul Gupta, the Head of Alliances at Digitate, joins the Intel on AI podcast to talk about how the ignio platform helps IT rapidly identify and remediate outages in minutes. He talks about how ignio blends AI, ML and advanced Deep learning to quickly resolve issues and preempt incidents when possible. He also emphasizes how ignio binds together disparate but interconnected business applications, processes, and the underlying infrastructure to drive smart decisions and perform actions autonomously. Atul explains how the Digitate team worked with Intel AI to use the Intel Distribution of OpenVINO toolkit to optimize some of their workloads and achieved incredible performance improvements. He also expresses how being a part of the Intel AI Builders program has been an excellent and productive experience for the Digitate team. To learn more, visit: digitate.com Visit Intel AI Builders at: builders.intel.com/ai

18 Marras 20198min

In this Intel on AI podcast episode: Creating, maintaining and deploying data models that allow businesses to gain actionable insights can require a lot of time and effort due to the vast amount of data to process. Also, machine learning optimization is a resource-intensive process and can be a challenge to accomplish in an industry where achieving fantastic results at speed is the final goal. Carlos Pazos, Product Marketing Manager at SparkCognition, joins the Intel on AI Podcast to show how the SparkCognition Darwin platform is transforming the way that data models are built in the enterprise world. He explains that automated machine learning (AutoML) enables even non-technical users to make use of science and drives organizations to scale the operationalization of ML models. The SparkCognition Darwin platform uses neuroevolution to custom build model architectures creating models in less time than traditional methods and enabling the rapid prototyping of scenarios and insights. Carlos also discusses how using Intel tools like the Intel Message Parsing Interface (Intel MPI) has allowed SparkCognition to provide greater performance to their end users. Also, how the Intel AI Builders program has provided an immense amount of development and marketing guidance to SparkCognition. To learn more, visit: sparkcognition.com Visit Intel AI Builders at: builders.intel.com/ai

31 Loka 201913min

In this Intel on AI podcast episode: Counting and identifying characteristics of crowds can provide organizations with a lot of valuable insights. Yet challenges like image distortion, density, and different camera angles can make analyzing images accurately very challenging. Ganes Kesari, Co-founder and Head of Analytics at Gramener, joins the Intel on AI podcast to discuss how Gramener has created a crowd counting solution that can overcome those challenges and produce a very rapid and accurate analysis of images. He talks about how Gramener has utilized this solution for several AI for good projects including a joint effort with Microsoft to count Antarctic penguin colonies. Ganes explains how their solution used convolutional neural networks (CNNs) using density-based estimations to deliver a more accurate penguin count than traditional manual counting methods. He also emphasized how benchmarking the solution on Intel AI technology and the Intel Optimization for PyTorch helped Gramener achieve comparable performance at a potentially lower computational cost. In addition to AI for good projects, Ganes also outlines how this same solution can also be utilized for other enterprise opportunities like drug discovery and how Gramener will continue to collaborate with Intel to provide better optimizations and performance for its customers. To learn more, visit: gramener.com Visit Intel AI Builders at: builders.intel.com/ai

27 Loka 201911min

In this Intel on AI podcast episode: Accurately answering clinical and billing questions by reading patient records, which can be hundreds of pages long, is a challenge even for human experts. While traditional rule-based or expression-matching techniques work for simple fields in templated documents, it’s harder to infer facts based on implied statements, the absence of certain statements, or a combination of other facts. David Talby, Acting CTO at John Snow Labs, joins us to talk about how the John Snow Labs Healthcare AI Platform and Spark NLP project are helping revolutionize the healthcare industry by addressing such workloads at a very high level of accuracy using state-of-the-art deep learning techniques applied to natural language processing (NLP). He discusses how NLP in healthcare is particularly challenging because clinical vocabulary and context can be very unique in comparison with other industries. David also explains how NLP is incredibly important for the healthcare industry because the data for many practical AI applications is trapped in text and NLP is needed to extract actionable insights. He illustrates how their solutions can run on-prem or in the cloud and highlights that John Snow Labs was recently recognized by CIO Applications magazine as its 2019 ‘AI Platform of the Year’ winner. To learn more, visit: johnsnowlabs.com Visit Intel AI Builders at: builders.intel.com/ai

20 Loka 201912min

In this Intel on AI podcast episode: Enterprises today are attempting to use Artificial Intelligence (AI) to tackle more and more complex challenges. Yet, many of the AI applications today are unable to cope with optimization and automation challenges in dynamic and complex environments like mobility, logistics, manufacturing and energy. Karim Beguir, Co-Founder & CEO at InstaDeep, joins the Intel on AI podcast to discuss how InstaDeep is helping their customers solve complex decision-making problems that would traditionally have been solved with existing algorithms but that can much better be served with AI and Machine Learning (ML). Some of the use cases that InstaDeep has tackled include working with large car companies to solve issues around ride-sharing and vehicle routing. Another example is how the company has helped customers in supply chain to optimize operations like container loading and bin packing. Karim explains how InstaDeep utilizes reinforcement learning which allows the algorithm to learn from itself and can model, simulate, and solve a problem without needing to have data from a customer in the first place. He talks about how collaborating with the Intel AI Builders program has enabled InstaDeep to develop solutions that provide better efficiency and savings to their customers. Karim also shares his vision for the future of an AI-first world and how InstaDeep is helping startups and companies around the world develop and utilize AI to improve their organizations and communities. To learn more, visit: instadeep.com Visit Intel AI Builders at: builders.intel.com/ai

11 Loka 201910min

In this Intel on AI podcast episode: Colorectal cancer is the one of the most common cancers worldwide, and in about half of patients it can spread to the liver and lead to greater complications. Traditional manual examinations of tumors are time-consuming and subjective for radiologists. Yet, using AI for this workload can be complicated because biomarkers, DNA, and genomic data are enormous and require lots of compute, storage and software to analyze. Saurabh Gupta, Director Advanced Analytics & Artificial Intelligence Product Management at SAS, joins us to talk about how SAS is working with the Amsterdam University Medical Center (UMC) on medical imaging analysis to identify if patients are candidates for surgery based upon the size and growth of lesions. He talks about how Amsterdam UMC using SAS AI-trained models running on Intel processors was able to analyze MRI images and with much greater accuracy identify patients who respond well to chemotherapy and become candidates for surgery. Saurabh highlights how SAS is not just simply a software vendor but also provides many tools that gives its customers a platform to do powerful AI analysis with the SAS software that they know and trust. He also talks about how important it is for SAS customers to be able to develop their solutions on Intel architecture because they already have the skills and architecture needed to address the workloads they have. To learn more, visit: SAS.com/ai Visit Intel AI Builders at: builders.intel.com/ai

3 Loka 201913min

In this Intel on AI podcast episode: There is a massive amount of insight to be gained by studying brick and mortar customer behavior. Yet capturing, labeling, and analyzing that data can be a very complicated and compute intensive activity. Dan Klein, Chief Data Officer at Valtech, joins the Intel on AI podcast to discuss how Valtech’s project VOID (Visual Object Identification) uses Intel® NUCs and the Intel developer stack to capture retail customer data like customer locations, face metadata, and apparel metadata to provide incredible real time insights for retail organizations. He talks about how powerful this technology can be to help retailers better analyze and cater to their customers in real-time but also how sensitive capturing individual information can be from a privacy and bias standpoint. Dan highlights how Valtech is proactively working to build their training data sets to include a diverse set of individuals of different genders and ethnic backgrounds to fight discrimination and bias in their analysis. He addresses the importance of ensuring individual privacy and fighting gender and racial bias in AI analytics and highlights that building a strong and diverse team at Valtech provides the foundation for this effort. To learn more, visit: valtech.com Visit Intel AI Builders at: builders.intel.com/ai

27 Syys 201913min