The Tesoro AI Podcast

This episode is action packed with advice on standing up a commercially viable AI product.  Our guest, Babak Hodjat (CTO of AI, Cognizant), is an incredibly successful AI founder with multiple successful exits.  Lucky for us, he also touches on some of his failures and challenges faced when building AI.  This is a must listen for anyone building AI products or teams internally.   If you are inspired by this episode, be sure to go check out Babak’s new book “The Konar and the Apple”.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.   Timestamps: 2:00  Babak’s experiences as a serial AI entrepreneur | Inspiring the natural language tech behind Siri 6:49  How Babak’s startup caught the interest of Marc Benioff and Sybase 8:52 A hedge fund innovator | Why AI has played a large role in stock market/trading related use cases 14:32  Babak’s role as CTO of AI at Cognizant | moving from research to commercial products 16:28  Measuring the impact/ROI of AI when feedback is not immediate 21:00  A starting point for making AI-driven decisions when there is little data 24:42  Unexpected challenges and opportunities when AI and humans work together 26:40  AI is not about big data, it’s about good data 31:40  How startups with no data can build a differentiated AI-native product 34:05  How your approach to delivering an AI prediction can be your sustainable competitive advantage 39:20  Identifying AI talent that can translate knowledge from academia to pragmatic problem solving 44:32  Check out Babak’s new book! The Konar and the Apple

Manufacturing is an industry that has taken early interest in the promise that artificial intelligence will deliver massive productivity gains.  Sectors such as autonomous vehicles are generating massive amounts of data that can be used to improve production and vehicle safety, reduce part recalls, and optimize costs using AI.   In this episode, we sit with Greta Cutulenco (Founder and CEO of Acerta Analytics).  Greta has been recognized as a member of Forbes 30 Under 30 in manufacturing.  Launching her startup in Waterloo, Canada, Greta and her team of 40 are enabling hyper focused on improving manufacturing quality using artificial intelligence.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.   Timestamps: 1:30  Early experience with autonomous vehicles + leveraging AI to improve manufacturing quality 4:00  Benefits of launching a startup in collaboration with a University: talent, data, trust 8:24  What is Acerta and why is it valuable? 10:00  How manufactures currently analyze machine generated data 14:05  How Acerta works with customer to collect data to drive analytics 19:47  Starting with zero data and finding partners to share data to build AI 24:05  Evolving from a University project to startup/commercial product 26:50  Establishing the initial team 28:30  Managing customer conversation in precision manufacturing 32:10  Adjusting to dynamic environments and data when operationalizing an AI model 37:30  Getting customers manufacturing operators to trust AI 43:20  The difference in fundraising as an AI startup vs traditional software

Google is the king of search.  But, is it the right search platform to use across all use cases?  Today’s guest, Leigh Fatzinger (Founder, Turbine Labs) explains why and how executive decision making requires a more contextually relevant solution.    SEO and ad spend can reduce the relevancy of information served up to executives while they are trying to make critical business decisions.  Turbine Labs has built a platform that removes those elements of search and remains hyper focused on providing the necessary information to help make business decisions quickly.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.   TimeStamps: 1:38  Leigh’s background and lead up to building Turbine Labs 3:08 Scaling Business Intelligence – Discovering AI as a tool to automate business processes 8:22  Example – Executive decision making | A once manual process automated by AI 12:26  Reducing mundane work for talented business analysts 16:25  Differentiation: Why Google search is not sufficient for executive decision making support 23:00  How Turbine Labs is used by executives 26:48  Acquiring enough data to train an ML algorithm 30:10  Turbine Labs’ approach to data labeling 35:22  Thinking through AI infrastructure as a non-technical AI Founder 38:21  Customer Conversations | What do customers care most about? 42:10  Explaining ROI through engagement metrics

Recently valued at $2 billion, Verbit is the latest AI-platform to reach unicorn status.  The startup has grown to $100mm in annual recurring revenue in only five years through its AI-driven transcription and captioning platform.    In this episode we sit with Tom Livne, Founder and CEO, to discuss several topics including how he leveraged the gig economy to build a more powerful AI product, growth through acquisition, how a vertical focus can drive better predictions, and much more.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.   Timestamps: 1:36  Verbit receives $2B valuation on 6x year over year growth 2:30  Tom’s journey from military special forces to law to tech entrepreneur 5:25  The Genesis of Verbit: discovering the challenges behind transcription editing 9:20 Use cases for Verbit’s transcription automation services 13:15  Recruiting a technical cofounder to be the brains behind the AI 16:06  How Verbit works 18:15  Determining a suitable accuracy rate for Verbit’s AI-driven predictions X20:15  How a focus on specific verticals impacted go-to-market and improved prediction capabilities 24:30  Building out a 35K freelance network of transcriptionists 27:35  Getting access to date to build an AI product 28:45  How Verbit acquired its first few customers + finding product market fit 34:12  Verbit raises $250mm Series E  and its plans for growth 35:53  A sneak peek into upcoming products 38:00  How Tom thinks about AI talent at Verbit 40:30 Growth via acquisition of mom and pop transcription services

Each bite of food we eat was prepared using a recipe with several ingredients with varying quantities depending on the food.  Apply that across a global food industry and you can understand the immense amount of data pouring out of this process alone.   Interestingly, the food industry still leverages dated processes for producing and delivering high quality food products.  Journey Foods has arrived as a data centric solutions for reducing supply chain costs while also recommending best fit ingredients en route to delivering tasty products.   Founder Bio: Founder and CEO Riana Lynn leads fast-growing software startup Journey Foods. The company is backed by top VCs in North America, Europe and Asia. Riana has served as a top consultant and VC to fortune 500 food and CPG companies. She's currently an angel investor and former Google Entrepreneur in Residence. She is one of just a few dozen black women that have raised more than a million dollars in venture funding.   Riana has been featured in Forbes, MIT 35 under 35, USA Today, Fast Company, CNBC, Wired, TechCrunch, Entrepreneur Magazine and more.   Riana Lynn is a Chicago native and Austin-based entrepreneur that enjoys growing fruit trees, writing film scripts, and exploring black culinary and architectural heritage sites around the world.   TimeStamps: 1:55  Riana’s journey as a serial entrepreneur 4:28  Finding a need for data insights as a solution for supply chain issues in food and package delivery 5:57 Packaged foods – a $3 trillion dollar industry still using archaic processes 7:13  Use case discovery: Where AI and data science could have the biggest economic impact 9:37  Reducing supply chain costs and improving product quality leveraging ingredient/product data 11:57  Reinventing archaic, inefficient processes in the food industry 15:02  Enabling sustainability in food product delivery and packaging 16:10 How startup food product companies are disrupting traditional players 19:22  Accessing GOOD data as an AI startup 21:17  Defining a data methodology at Journey Foods 24:47  Deciding to build as a solo founder 26:42  Initial customer conversations and go-to-market with an MVP 28:50 Moving from SMB to enterprise clients 31:26  Journey Foods fundraising experiences 36:17  Riana’s vision for accelerating Journey Foods growth in South America and Africa 37:52  Journey Fods’ ideal venture partners

Manufacturing and design firms continue to stumble upon novel use cases for artificial intelligence.  Today we learn more about how 3D imaging data is reducing supply chain complexity for manufacturing parts.   Startups also have an interesting challenge in acquiring the data necessary to train machine learning models.  In this episode with Paul Powers, Founder and CEO of Physna, we dive into what it takes to build an AI startup.  We also discuss his early career in the legal field…in Germany.   Timestamps: 1:39  Launching a career in law in Germany 3:13  Where the idea for Physna originated 7:04  Funding Physna before raising venture capital 8:44  The evolution of Physna | Pivoting to a higher value use case in engineering productivity 16:32  How artificial intelligence is used to understand relationships in 3D images 19:45  Leveraging unique data sets to deliver customized, more accurate predictions 23:53  Acquiring the initial data a startup needs to train AI algorithms 28:25 Establishing data partnerships with the enterprise and public organizations 37:28  How much do customers care that AI is driving your product? 39:00  Perspectives on fundraising 45:00  Thangs.com – See what Physna can do…for free

There is always room for AI that improves the quality of human life.  In this case, we examine an artificial intelligence solution that is actually saving lives.  For stroke victims, the length of the process from scan to treatment is often the difference between life and death.  Today, many hospitals deal with preventable death due to processes that delay wait times for treatment of stroke patients.   In this episode, I sit with David Golan (Co-Founder and CTO of Viz.ai), to discuss various topics including AI in healthcare, data acquisition from startups, how AI forces you to reevaluate workflows, and much more.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.      01:39  How a stroke mimic inspired David’s to build an AI to reduce healthcare system complexity 4:57  How strokes are caused and why patients experience delays in treatment 9:15 The genesis of Viz.ai – leveraging automation and deep learning to interpret scans 10:55  Working with physicians to determine best UI/UX 15:32  Determining the accuracy level required to deliver a high performing AI solution 17:38  Acquiring the data to train an ML model in healthcare 19:28  Solving challenges around data annotation 21:07  Great AI research vs great AI products 27:39  The importance of understanding customer workflow when developing AI solutions 29:41  Finding early adopters for the product 32:28 Favorite tech tool at Viz.ai 35:00  The people who have inspired David?

AI is often referred to as a “black box” as AI practitioners struggle to explain why a model generates a given prediction.  As AI is being deployed into the real world, organizations are not only tasked with producing high performance models but also with avoiding AI models that are bias (and have a negative impact on society).   In this episode, I am joined by Will Uppington, Founder and CEO of TruEra.  He and his cofounders have set out to enable deployments of high-performance AI solutions that are transparent and build trust.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, data labeling, and AI product development.   2:00  Before TruEra: Discovering the challenges in AI quality, model explainability and monitoring 5:00 Finding co-founders out of academia 6:00  Use case selection in the enterprise 8:00  Where and how to get the most business value out of AI? 12:03  The advantages of iteration when in comes to machine learning 13:45  Building a team of founders focused on societal impact of AI 16:34  The importance of explainability in AI 19:16  Speeding up the discovery and debugging of bias AI models 22:20  Engaging with the TruEra platform – how it works 26:47 TruEra’s value proposition explained 34:06 Most helpful tech tools at TruEra 35:00  Who has inspired Will most?  

Legal fees – the line item in your budget you hate to see, but also the critical service that keeps your business afloat.  Most executives haven’t the slightest idea of what legal services should cost and how what they are charged compares to the market rates.  Simply put, transparency in legal spend does not exist – until now.  Bodhala is leveraging AI to demystify and recommend legal services based on talent requirements and internal budget.   In this episode, I sit with Raj Goyle, Co-Founder and CEO of Bodhala.  We discuss his background as a Congressman, why legal spend is an important problem to solve, developing a data strategy as a startup, and much much more.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.

AI is often thought of as the tech that is going to replace human workers.  Less often do we see opportunities for AI to improve and protect human life.  Active shooter threats have begun to become a more common media headline.  This issue also plays itself out in the commercial sector as disgruntled employees create unsafe environments in the workforce.  Security cameras are common and spread throughout our cities, but with AI we can now use those cameras to alert authorities to deadly threats, reducing response times exponentially.   In this episode, I sit with Mike Lahiff, CEO and Chairmain of ZeroEyes.  We discuss his career as a Navy SEAL, finding the right technical talent to build AI products, navigating the data journey, launching a public safety product, and much much more.   If your company is looking to scale it’s AI initiatives, head over to Tesoro AI (www.tesoroai.com) – experts in AI strategy, staff augmentation, and AI product development.   Timestamps: 1:38  Mike Lahiff – Navy SEAL, private equity investor, AI founder 5:03  The origins of ZeroEyes – Solving the active shooter problem in the US 6:32  Finding talent to build the initial AI product 8:25  The data journey – inaccurate predictions that inspired better data collection processes 10:09  Acquiring early partners to demo and improve product 11:03  Building an AI product as a non-technical Founder “selling the mission” 14:43  When did it make sense to bring on technical AI talent as permanent team members 16:51  Building in-house vs outsourcing software development and data collection/prep 21:00  Building and selecting tools to speed data annotation process 23:10  Testing the product with the local police force | 50% reduction in response time 26:49  Expanding to commercial organizations 28:56  Self-funding the first $500K, then raising formal VC 30:33  Greatest challenge to building AI-driven software 32:12  Favorite tech tool: Slack, Label Box 33:17  Who inspires Mike