CrowdFlower, a well-known microtask crowdsourcing business founded in 2007, announced this week that it had raised $10 million in a Series D round. That brings the total funding for the company to $37 million. The round was led by Microsoft Ventures, with Canvas Ventures and Trinity Ventures in tow. The investment will primarily be used to advance the new CrowdFlower solution offering called “CrowdFlower AI,” which provides CrowdFlower’s customers the ability of to economically utilize artificial intelligence (AI) and machine learning.
In its earlier years, CrowdFlower provided straightforward crowd microtask solutions. Customers could launch projects, such as data collection and categorization and sentiment analysis. The platforms would distribute microtasks to the CrowdFlower crowd, aggregate the results and then provide a processed data set back to the customer. Through the platform, CrowdFlower also imposed quality control of the crowd and its performance and results.
Later on, CrowdFlower also began to focus on serving data scientists and companies that were using their own AI and machine learning algorithms to build their own predictive models. One critical step in this process is the “training” of the algorithms based on large quantities of data that must be properly cleansed, classified and otherwise structured. CrowdFlower addressed this need with a new offering, “Training Data.”
Then, in October of 2015, CrowdFlower quietly launched “CrowdFlower AI.” The CrowdFlower platform was augmented with its own AI/machine learning capabilities, which allowed customers to build their own predictive models for their specific use cases, but without having to invest in their own systems. This makes AI/machine learning-based modeling more economical and accessible to a larger market and applicable to a larger number of use cases that could be addressed previously. Once might also think of this as AI/machine learning-as-a-Service.
“We wanted to bring the economics of applying AI and machine learning within the reach of every business,” CrowdFlower COO Robin Bordoli wrote in a recent blog post. Bordoli also said that increasingly data science teams were using the platform “to create the training data they need to feed their machine learning algorithms. Given our front row seat it made sense for us to make machine learning a commercially viable proposition for our customers.”
“We saw an underserved part of the market, Bordoli added. “Namely the millions of million-dollar business problems such as support ticket categorization. These problems aren’t worth a billion dollars to solve because the way in which each company wants to solve these problems is custom to them.”
At this time, in addition to its prior offerings, CrowdFlower enables three types of capabilities to customers: training data, machine learning and what it calls “human-in-the-loop.” People are not only used to create training data, but also to assist in the machine learning to improve predictive accuracy.
Two observations about these new offerings and the recent investment:
First, the strategic expansion of the potential addressable market through the three capabilities mentioned above apparently sits well with investors. Nagraj Kashyap, corporate vice president at Microsoft Ventures, was quoted as saying in the recent press release, “At Microsoft, we’re looking to create experiences for people and businesses where technology intelligently supports what they’re doing. CrowdFlower’s approach — combining human and machine intelligence to solve all types of unstructured data problems — aligns with that effort. We look forward to supporting them in their next phase of growth in the broader machine learning and AI market.”
Second, the evolution of a microtask crowdsourcing work intermediation platform like CrowdFlower is consistent with trends we have been observing in the space. We recently wrote about this, noting “Increasingly, online work intermediation platforms are not just about matching and workflows — data analytics, AI and machine learning and other algorithmics are being added to platform stacks and becoming a part of the (human) work intermediation (and value-creation) process.”
In other words, we are seeing new platform models that add AI/machine learning capabilities and keep humans around in roles that are unique to them and add value to the whole process (racing with the machine, not against, as proposed by Erik Brynjolfsson and Andrew McAfee in Racing Against the Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity and Irreversibly Transforming Employment and the Economy).
And the relevance of all this to services procurement professionals? We wrote in “Crowdsourcing and Cognitive Computing: Are You Ready for the Future of Work?,” procurement will still be tasked to enable and manage new platform-based work/service intermediaries (“suppliers”) — with the usual objectives in mind (cost, risk, performance). But we are talking about “suppliers” that are radically different from the suppliers we have been dealing with for 10 to 20 years.