Why I’m Joining integrate.ai

I am excited to announce that I’m joining integrate.ai as Chief Revenue Officer.  I spent several months consulting for them as I learned more about this fascinating AI space and the market for data science tools.  I wrote some thoughts on what I’ve learned and why I’m so optimistic about where we’re headed.  

‍

Federated Learning: 

This company has been around for several years working on a technology called federated learning.  Normally, when you want to train a machine learning model on data, you have to move all the data into one place first and then you can perform data science tasks on it.  But with federated learning, you’re able to keep the data where it is, and essentially move the model to the data.  This comes in pretty handy in certain situations where the data is too sensitive to move, like in certain healthcare contexts, cross-border situations, or even voice assistants like Siri or Alexa. 

‍

Signal Hunting:

We’re using the technology to help companies with what we call “signal hunting.”  Companies like banks and insurance carriers building models for pricing and underwriting, as well as pharmaceutical companies building models for small molecule discovery, all need data to make their models more accurate.  But the process of sorting out the signal from the noise is so painstaking, that they often don’t even attempt it unless they’re already confident the data will help.  But one of the key benefits of AI is that it can find connections between data that humans can’t.  These companies are amassing large data science teams, but they are essentially blocked from free experimentation because they don’t have access to enough data, and getting access is too slow and difficult.

‍

A Dressing Room for Data:

integrate.ai has built a platform that serves like a dressing room for data.  Imagine someone going to a department store to buy a $500,000 dress, and they can’t try it on first.  Maybe they get a swatch of the fabric (e.g. sample data), but they can’t even look at the whole dress.  This is what it’s like to evaluate data right now.  A department store would never say, give us your shoes and jewelry, and we’ll match them up to the dress and then give you a report about how well the clothes work together.  And yet, that’s the current state of the art in the world of buying and selling data.  Not to mention that the process costs you $100,000 and takes 3 months. 

‍

A Playground for Data Science Experimentation: 

The integrate.ai dressing room, powered by federated learning, is like having the entire department store in your own closet at home.  Data scientists can test and validate their models against any 3rd party data, without having to hand over their own valuable data, and the department store doesn’t have to hand over their data.  They can even fully train a model on the data to see exactly what impact it will make.  The platform restricts the data buyer from using their model in production until they’ve paid for the data.  Kind of like those little alarm buzzers that they take off the dress at the cash register once you pay.  The end result is a playground for data scientists to freely experiment with data, without having to know beforehand that the data will be valuable, and without having to go through the long and arduous testing process in place today.  We believe this will create an exponential increase in AI innovation among these data science teams within the enterprise.

‍

The Four Phases of Data Collaboration in AI: 

When OpenAI released ChatGPT - they trained it on all the world’s public data - the entire free Internet.  This is the open data phase.  The next wave is already happening where enterprises want to combine that model with their own in-house data to make chatbots that know about their internal workings - we can call this the 1st party data phase.  But that won’t be sufficient - most predictive models already in production within the enterprise use 3rd party data, and GenAI models won’t be different in that respect.  The next phase will be to improve those models with 3rd party data, where all these enterprises will go signal hunting, and the first place they’ll look are all the data providers who sell data today.  These include companies like credit bureaus and data analytics firms.  This is the 3rd party data phase.  Then the last phase will be the dark data phase, where the enterprises will burn through all the data out there that’s for sale, and will go looking for data that’s not for sale.  At integrate.ai, we’re already enabling the 3rd party data phase, and we believe that the dark data phase will only be possible with a platform like ours that has the rails for data collaboration that doesn’t require enterprises to be passing sensitive data back and forth.

This is an incredibly ambitious project, with some killer technology and amazing people behind it.  I couldn’t be more excited to join forces and help them bring this to market.  

‍

About Bob

‍Bob is the Chief Revenue Officer at integrate.ai.  Bob is a GTM executive with over 20 years of experience in Fintech, Payments, E-commerce, Adtech, Enterprise SaaS and B2C.

Prior to integrate.ai, Bob was the Chief Business Officer at the Fintech payments firm Bolt, where he helped grow the company by 20x during his tenure through his leadership of partnerships, channel sales, business development, alliances, corporate strategy, M&A, and as GM of the social commerce business.  Prior to Bolt, Bob was VP of Sales and Business Development at SoFi where he managed several teams responsible for acquisition for a large portion of the lending portfolio, and helped launch the company’s new banking, investing and mortgage products through strategic partnerships.  

Prior to his fintech experience, Bob was the CEO of Manifest Commerce, a social ad-tech company that was acquired by Rakuten.  Post-acquisiton, Bob served as the SVP of Sales for Rakuten’s Advertising division.  Before Manifest, Bob was the VP of Business Development at Digg where he created DiggAds, the first native ad format for social platforms..  

Bob is a Lead Mentor with StartX, Stanford's startup accelerator program, and also volunteers his time as a pro bono attorney for LegalAid.  For fun, Bob surfs, rides bikes, cooks, and plays guitar and sings in a dad-band.