Resources

Newsroom

Back to Industry Trends & Insights

The Latest Advancements in AI Presented at VentureBeat Transform 2019

Sample Venturebeat Slide

AI Foundry Joins Top Industry Leaders Presenting AI Solutions at Industry Conference

July 17, 2019

By Steve Butler, Founder & President, AI Foundry

AI Foundry was selected as one of 15 companies to present at Venture Beat Transform, a top technology industry event held last week in San Francisco. Jeff Goodrich joined me to present a demo of our new mortgage document process model, which was announced July 10th. We spoke to a standing-room-only crowd at the event, with some of the top executives in technology, science, and venture capital among the 1,200 in attendance. We were honored to be presenting at such a prestigious, high-demand event.

After we left the stage, we spent the remainder of the conference listening to other presentations and having conversations with attendees and presenters. It was a great opportunity to network and see what other AI companies are up to. I wanted to give you my thoughts and perspective from the two days that I spent at VentureBeat Transform.

AI and Big Brand Companies

The theme for this year’s event was AI, and we saw many interesting applications of AI across many industries, as well as several big-brand companies. For example, there was an entire two-day track that focused just on computer vision, and included presentations from companies like Pinterest, Royal Caribbean Cruise Line, Walmart, Wayfair and Zillow. From these presentations, you could see that real progress is being made in computer vision technology.

Common Challenges in AI

Across many of the presentations and discussions, we heard about number of common challenges with AI. And, these are some of the same challenges we face every day. They include:

  • Acquiring and annotating imagery for training
  • Setting up the data and machine learning pipeline
  • Handling the uncertainty and inaccuracies in the model
  • Managing concept drift when real-world data changes
  • Effectively handling the human collaboration with AI

Many Positives, and Key Takeaways

First, we were pleased to have some great feedback. Several people came up after our presentation and said “you guys have a real business with real customers and revenue. You are way ahead of most of the other firms that presented.”

My overall takeaway from the first day of the show was that AI is a very active area and we have a very novel and impressive application.

Here are a few specific items that stood out to me:

  • Computer Vision – I could not find anyone else using computer vision on documents. Many are doing object recognition in areas such as shopping or for counting people in different venues, or autonomous cars. From this, I learned that AI Foundry really has a unique application, which is very exciting!
  • Accuracy in AI – the CV accuracies that most people talked about was 70-80% for applications such as shopping. No one made any claims even close to the 90-95% accuracy that AI Foundry has reached.
  • Data Prep and QA – no one has any “secret sauce” when it comes to data preparation or quality assurance. I found that other companies hired “third party” resources to do manual curation and labeling and post model quality assurance.
  • Self-Learning is on the Rise – Self-learning from exception handling appears to be available only based on the degree of exception scenario. But it is getting smarter and this area has great potential.

Amazing Potential for AI

After two days at VentureBeat Transform I was feeling that AI’s potential is limitless. There were discussions and demos on AI applications in virtually all areas of life, including:

  • Mobile-based biometrics (very cool)
  • Technology that makes current facial recognition no longer work, but replaces it with a better solution
  • AI that makes cruise ships a better experience
  • Neural chips that claim superiority over GPUs for AI applications
  • AI that personalizes the experience for Airbnb users,
  • All kinds of new shopping technology (bots, vision-based search, NLP-based search), and so much more.

It is clear that all the companies believe that getting close to the customer is the best way to be successful, as trials with customers and real data were key parts of presentations. Belief in the DevOps/CI/CD model was strong. I saw very little development in isolation, or the Field of Dreams “build it and they will come” strategy.

In all, AI Foundry stacks up very well with all the firms I saw, including the largest companies. We have a very modern architecture, tons of open-source technology, a hybrid tool stack, and a Devops CI/CD collaboration strategy with early customers. In addition, we are using AI development best practices, and it is exciting that we are leaders in the process and Dev area.

Also, it’s important to note that many of the startup firms at the conference are looking for first revenue and have no evidence of ROI or value yet. AI Foundry’s customers, revenue and pipeline really set us apart from the pack. In all, the pieces are in place for us to create a major company, based on what I observed the past few days.

The conference was a great experience and gave me a lot of hope for the future of AI and reinforcement of the great work that we are doing at AI Foundry.


Watch VentureBeat Presentation +  

Read Press Release + 

-------------------------------------------------------------------------

Steve Butler 
Founder & President

Steve leads AI Foundry, which automates and streamlines enterprise business operations. Under his leadership, AI Foundry creates software solutions that makes data “actionable”. Steve’s leadership includes CEO positions for high-performance data analysis, enterprise software, system management and CAE companies.

Connect with Steve on LinkedIn