By Robin Bordoli, June 12, 2017

Making AI Work in the Real World

Today we announced our recent $20M financing led by Industry Ventures, with new investor Salesforce Ventures participating alongside existing investors Canvas Ventures, Freestyle Capital, Microsoft Ventures, Promus Ventures, and Trinity Ventures.  It’s only been a year since we announced our $10M financing led by Microsoft Ventures.  This post explores why we’ve raised another round so quickly and what we plan to do with the capital.

So first, why did we raise $20M capital so soon after raising $10M?

It’s a simple answer.  We’re at the epicenter of a massive exploding market and our new and existing investors are voting with their dollars that CrowdFlower is well positioned to respond to the market demand.  That market is the application of AI and machine learning inside the enterprise. As a company we’re focused on the mission of making AI work in the real world.  The real world is messy and complicated, but that’s what makes it challenging and worthwhile.  As JFK said “We choose to go to the Moon in this decade and do the other things, not because they are easy, but because they are hard”.

So what’s the evidence that the market is massive and exploding?  

The evidence is there for all to see.  According to LinkedIn there are almost over 300,000 data scientists or machine learning engineers in North America.  We’ve been helping these data science and machine learning teams take the first step on their AI journey converting their unstructured text and images data into high quality large scale training data on our human-in-the-loop platform.  We’ve done this across a broad range of use cases – autonomous vehicles, facial recognition, aerial and satellite imagery, medical image labelling, consumer product identification in retail environments, natural language processing, text classification, content categorization, product categorization and search relevance.  AI is reaching into every industry, every market, every corner of human endeavor.

In addition to the rise in the data science and machine learning job function, we’re also seeing major investments from Amazon, Google, IBM, Microsoft, and Salesforce in bringing machine learning into the cloud so every business can access the power of machine learning without having to build their own stack.  This means that companies can now start to apply AI to their business processes for hundreds of thousands of dollars rather than tens of millions of dollars.  This means AI is now within the economic reach of upwards of 20,000 companies in the U.S.

Second, how are we going to invest the $20M?

We’ve been increasing our levels of product and market investment to serve this demand.  We recently launched our Image Annotation for Computer Vision solution with bounding box, dots, polygons and semantic segmentation capabilities. We’re helping researchers, academics and social entrepreneurs apply AI to society’s big challenges with the $1M “AI for Everyone” Challenge.  We just hosted 700 people at Train AI last month focused on the challenge of machine learning in a human world.  We brought together the AI pioneers from companies such as Adobe, Amazon, Google, Kaggle, HERE, LinkedIn, Microsoft, Planet Labs, Reddit, Salesforce, Spotify and others so companies could learn from them.  If you missed it, you can check out the presentations here.

But we need to do more so we raised this capital to invest in the 3 Ps – people, product and partners.

  • People.  We’re an 80 person startup in the heart of the Mission in San Francisco with a big audacious mission of making AI work in the real world.  We’re investing across the business in hiring people who are passionate about this mission.  We’re hiring engineers, product managers, and machine learning experts.  We’re hiring project managers in contributor operations which is the heart of our how our platform operates in applying human intelligence.  We’re hiring marketing and sales professionals to help us find and win new customers.  We’re hiring in customer success to help our customers get the most out of our platform and achieve the results they need.  We’re hiring in finance and HR to help our organization scale.  Check out all our job openings here.


  • Product.  Our human-in-the-loop AI platform has more scale than any other comparable platform.  Our platform has executed over 2 billion judgments where a human has annotated an image or categorized a product or determined a sentiment.  But we don’t want to stand still.  We want to accelerate our product investment so that our customers can get ever larger training data sets at the quality levels and price points that meet their business needs.  We want to make it easier for CrowdFlower to integrate with machine learning vendors. We want to make supervised learning easier and more practical so data scientists can focus training data collection where it’s most needed.  Our CrowdFlower AI powered by Microsoft Azure Machine Learning solution is representative of our first step in this direction, but we can do so much more.  Stay tuned for more on new product capabilities in the coming months.


  • Partners.  Having Salesforce and Microsoft as partners in our business is an important step in our continuing journey on becoming a large independent company that is an essential part of the AI ecosystem.  We’re a natural complement to the major technology players Amazon, Google, IBM, Microsoft, and Salesforce who are providing the algorithms in the cloud.  We provide the training data and the human-in-the-loop workflows which teach, test and course correct the algorithms.  The intertwined nature of algorithms and training data has been well captured by our VP Marketing Randi Barshack in our billboards.


Finally I want to say a big thank you to our customers.  We’ve been fortunate to have hundreds of companies invest their time and money in CrowdFlower.  We recently had the chance to host an appreciation dinner for some of them. We have only reached this point in our journey because of the faith you’ve placed in us.  We take that trust seriously and we recommit ourselves to helping you make AI work for your business.  We won’t always get it right the first time, but we will always be honest and humble partners who persevere and in JFK’s words will give the “best of our energies and skills, because that challenge is one that we are willing to accept, one we are unwilling to postpone, and one we intend to win”.

Onwards and upwards.