Episode 120: Your AI Insurance Appraiser Has Arrived with Tractable CEO, Alex Dalyac


This post is by David Poole from Georgian Partners

The process for making a car insurance claim is slow and cumbersome. Waiting as an appraiser with a clipboard goes over the damage, wondering how their biases and human fallibility might affect the outcome. That painstaking process includes data collection though – and it turns out that dataset is the perfect training ground for AI.

Alex Dalyac is our guest on this episode of the Georgian Impact Podcast. He’s the Co-founder and CEO of Tractable AI. Until recently humans had the edge over AI when it comes to image classification tasks – but the scales have now tipped in the computer’s favor. Tractable is leveraging that fact to revolutionize the way damage from automotive accidents and natural disasters is assessed.

You’ll hear about:

  • How accident and disaster damage appraisals could be 10x faster using AI.
  • Why Tractable chose to pivot their AI’s strengths from plastic pipes to the insurance industry.
  • How Alex and his team convinced competing insurance companies to pool their data – and how they keep that data safe.
  • The challenges of selling in such a consolidated industry.
  • Tractable’s approach to improving trust and transparency.

Listen to every episode:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

Who is Alex Daylac?

Alex Dalyac is the Co-founder and CEO of Tractable AI , an artificial intelligence company specialized in visual tasks for traditional industries. The company’s current focus is insurance and automotive, where its AI predicts the cost to repair a vehicle based on photos of the damage. Its products are used by leading insurers in Europe and North America. Tractable was spun off from Alex’s research at Imperial College London, where he led the Computing department’s first industrial application of deep learning. Prior to research, Alex was at hedge fund Quant.

The post Episode 120: Your AI Insurance Appraiser Has Arrived with Tractable CEO, Alex Dalyac appeared first on Georgian Partners.

Episode 119: You Need a Data Strategy with Immuta’s Dan Wu


This post is by David Poole from Georgian Partners

A solid data strategy can prevent your company from running aground and turning a huge opportunity into a horrible mess.

Dan Wu is our guest on this episode of the Georgian Impact Podcast. Dan is a superstar commentator in the privacy and data governance space. He’s leveraging his Ph.D. in Sociology and Social Policy and his law degree to help protect people and their data. Dan believes that the best way to do that is through data strategies formed by cross-functional teams that include input from governance, analytics, marketing and product departments.

You’ll hear about:

  • What we can learn from the botched launch of the Apple Credit Card
  • Why every company needs a data strategy
  • How regulation, like the Algorithmic Transparency Act, could add protections for consumers and accountability for business
  • Offensive vs. defensive data strategy – HBR Article
  • Where responsibility for inaction leading to data breaches should lie
  • Data risks businesses face, including biased algorithms, sharing data with the wrong people, 3rd party data breaches, insider incidents, and technical mistakes
  • Why data ethics need to go beyond what’s strictly legal in order to establish and maintain trust.
  • AI Now’s 2019 report that touches on ethical inequality risk factors in AI

Resources:

Listen to every episode:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

Who is Dan Wu?

Dan Wu is the Privacy Counsel & Legal Engineer at Immuta, a leading automated data governance platform for analytics. He writes about purposeful data strategy on TechCrunch and LinkedIn. He holds a J.D. & Ph.D. from Harvard University.

The post Episode 119: You Need a Data Strategy with Immuta’s Dan Wu appeared first on Georgian Partners.

Episode 118: The Business Case for Deep Fakes with Descript’s Kundan Kumar


This post is by David Poole from Georgian Partners

Deep Fakes are incredibly realistic impersonations that blur the line between truth and fiction. So what happens when the tech to make them is available to everyone? We’re about to find out.

Kundan Kumar is our guest on this episode of the Georgian Impact Podcast. He is co-founder of Lyrebird AI, and now heads up research at Descript. Descript is making it vastly easier for anyone to manipulate audio, even to the point of inserting words that were never actually said. This brings up obvious ethical, trust, and security questions. Fortunately that’s something Kundan and the company are putting a lot of thought into.

You’ll hear about:

  • How anyone can create a voice-double, and the ethical questions that raises.
  • Automating common audio editing tasks like removing ums and ah, and bleeping swear words.
  • Georgian’s podcast production workflow – and how we manipulate what you hear.
  • How these new technologies are, in many ways, evolutions of exiting machine-assistance features like auto-correct, and Google’s Smart Compose
  • Maintaining trust when trickery is effortless and skepticism is ubiquitous.

 

Listen to every episode:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

 

Who is Kundan Kumar?

Kundan Kumar is Research Lead at Descript and co-founder of Lyrebird AI. He is a PHd student at Mila, Quebec’s artificial intelligence institute, where he works on generative models for sequences e.g. speech and music.

The post Episode 118: The Business Case for Deep Fakes with Descript’s Kundan Kumar appeared first on Georgian Partners.

Episode 117: Leading From Home


This post is by David Poole from Georgian Partners

Working from home isn’t new, but today an unprecedented number of people are “telecommuting” for the first time – whether they want to or not.

Our guest on this special edition of the Georgian Impact Podcast is Ben Wilde, our VP of Business Development & Marketing. Ben manages a team in Toronto from his home office in New Zealand. He’s been leading from afar for years, and shares his tips and insights with all those adapting to new ways of working during the global pandemic and beyond.

You’ll hear about:

  • How work may change forever A.C. (After Coronavirus)
  • Informal management of remote teams when you can’t pop by their desks
  • Etiquette for video meetings and time-shifted communication
  • Creating boundaries and structure when working from home
  • Building trust through casual conversations and social bonding
  • Being conscious about your approach to remote work

 

Listen to every episode:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

Who is Ben Wilde?

Ben Wilde is Vice President Business Development & Marketing at Georgian Partners and a member of the Georgian Impact team. He provides input to overall strategy including product management and product marketing.

Ben has more than 20  years of experience in the software industry, and has held senior product management and strategy roles at IBM, Informix, and a range of early and growth-stage startups. Ben holds a BTech (Hons) in product development from Massey University, New Zealand.

The post Episode 117: Leading From Home appeared first on Georgian Partners.

Building Effective Machine Learning Teams


This post is by David Poole from Georgian Partners

The post Building Effective Machine Learning Teams appeared first on Georgian Partners.

Building ML Teams Resources


This post is by David Poole from Georgian Partners

Resources to Build
an ML Strategy
and Team

We’ve pulled together twelve top resources to help you set the vision, prioritize opportunities, and hire a remarkable team. To get started with AI, there are critical steps that must be taken to ensure success.

Understanding the feasibility of the opportunity comes first, and having the right team is next. If you are considering AI for your business, these resources are an excellent place to start.

Strategy Resources

1. Georgian Partners Principles of Applied AI

This white paper introduces our ten principles of applied AI and lays out a detailed framework for getting started. The paper shows how you can avoid common mistakes and provides you with a method of measuring your progress. You’ll gain a better understanding of what business AI problems can solve and what it would take for you to implement a solution.

2. Andrew Ng’s AI Transformation Playbook

Andrew Ng’s AI Transformation Playbook delivers insights gleaned from his days leading the Google Brain team as well as the Baidu AI Group. By following the Playbook, Ng says that any AI-focused buisiness can gain the same kind of momentum and success he details here. The Playbook has a strong focus on building an in-house AI team as well as how companies can leverage it to their best advantage.

3. AI Canvas Presentation

The AI Canvas is a methodology that helps you build business cases for AI projects. It will prove your concept and Continue reading Building ML Teams Resources

Episode 113: Taking Control of our Personal Data

How can you tell the difference between a company that resells your data and one that treats it with care? How do you control where it ends up? In this episode of the Georgian Impact Podcast, Jon Prial is joined by Lisa Levasseur, founder of the Me2BAlliance. They discuss a future where consumers are more educated and companies adhere voluntarily to standards to earn the trust of their customers.

You’ll hear about: 

  • Developing a standard for personal data management
  • How to develop standards for products and services that are trustworthy
  • Why disclosing business models lays the foundation for trust
  • How consumers can manage their data

Listen to every episode:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

Who is Lisa Levasseur?

Lisa Levasseur is the founder of the Me2BAlliance. Lisa has been a software professional for over 30 years, with her formative career years spent Continue reading Episode 113: Taking Control of our Personal Data

Episode 112: Designing for Humans and Machines

Human-centered design has helped to make truly great products, by designing the product so that the user has a great experience and perceives value. Now, to be successful, we have to design products that consider intelligent machines as users and optimize the interactions they have with human users.

In this episode of the Georgian Impact Podcast, Jon Prial and Lindsay Ellerby, Senior Designer at Normative, discuss the concept of centaur design – designing solutions for a world of AI-human hybrid systems.

You’ll hear about: 

  • What EQ looks like in machines
  • How to think about the user in an age of intelligent machines
  • How to create positive human-machine interactions
  • What centaur design is and how it will impact the future of design

 

Listen to every episode:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

Who is Lindsay Ellerby?

Lindsay Ellerby is a designer who excels Continue reading Episode 112: Designing for Humans and Machines

Episode 110: Data Trust by Design with Nathan Kinch

On previous episodes of the Georgian Impact Podcast, we’ve talked about Privacy by Design with Ann Cavoukian and Privacy as Trust with Ari Ezra Waldman. In this episode, we complete the trilogy by discussing Data Trust by Design with Nathan Kinch. 

Trust can be a difficult topic to pin down and make actionable. Nathan and Jon discuss how Greater than X has built a toolkit called Data Trust by Design to provide a systematic method through which organizations can design for trust and actually be worthy of it. 

You’ll Hear About:

  • Why customers are willing to share data when they trust the company and product
  • How you can sell more on a positive message of trust
  • Which metrics trust impacts
  • How brands with trust outperform the competition


Subscribe:
iTunes | Spotify | Google Play | SoundCloud | Stitcher | RSS

Who is Nathan Kinch?

Nathan Kinch is founder Continue reading Episode 110: Data Trust by Design with Nathan Kinch