Loading organizations...
Key people at CrowdFlower.
Based in San Francisco, California, CrowdFlower provided a crowdsourcing SaaS platform that combined human labor with machine learning to manage distributed workers for data labeling and microtasks. The company served both self-service users and enterprise clients requiring artificial intelligence data preparation, including customers such as Zvents and O'Reilly Media. Over its operational history, the enterprise secured multiple rounds of financing, including a $12.5 million Series C in 2014, a $10 million Series D in 2016, and a subsequent $20 million funding round in 2017. These capital raises were backed by prominent institutional investors, with participation from Microsoft Ventures, Salesforce Ventures, and Trinity Ventures. The organization eventually rebranded as Figure Eight and expanded its international footprint before being acquired and ceasing independent operations in 2020. CrowdFlower was originally founded in December 2007 by Lukas Biewald and Chris Van Pelt.
CrowdFlower, founded in 2007 as Dolores Labs, developed a platform combining crowdsourced human labor with machine learning to create high-quality training data for AI models, addressing the challenge of labeling unstructured data like text, images, audio, and video.[1][4][6] It served data science teams at major companies including Google, Facebook, Autodesk, and financial services firms, enabling applications in autonomous vehicles, natural language processing, search relevance, and chatbots; the company raised about $58 million, achieved strong revenue growth (nearly tripling in one year pre-2017), and rebranded to Figure Eight in 2018 before being acquired by Appen in March 2019.[2][3][4]
CrowdFlower was founded in 2007 by Lukas Biewald and Chris Van Pelt, initially as Dolores Labs in San Francisco's Mission District, after they identified a need for human workers to handle simple, non-automatable tasks.[1][3] Biewald, drawing from his time at Yahoo, experimented with Amazon Mechanical Turk on facial assessments via Facestat, collecting 20 million data points in three months and attracting early clients like Zvents and O'Reilly Media; the company quickly outgrew its space, rebranded to CrowdFlower in 2009, and launched officially at TechCrunch50.[1][3] Pivotal early support came from Travis Kalanick (pre-Uber), who invested $50,000 and aided pitches, alongside angels like Dave McClure and Jeff Hammerbacher; by 2009, it secured a $1.15 million seed round amid the financial crisis when machine learning was undervalued.[3]
CrowdFlower rode the early wave of machine learning and AI data needs, emerging in 2007 when ML was niche and investor-skeptical, but timed perfectly for the AI boom by solving the "human side" of training data bottlenecks that algorithms alone couldn't handle.[3][7] Market forces like exploding unstructured data volumes and the rise of deep learning favored its scalable crowdsourcing model, powering ecosystems in autonomous driving, NLP, and fintech while influencing standards through open datasets and human-AI hybrid approaches.[1][4][6] It accelerated AI adoption for startups and giants by democratizing data labeling, though competitors in reinforcement and transfer learning loomed as potential disruptors.[7]
Post-2019 acquisition by Appen, CrowdFlower's legacy as Figure Eight endures within a larger AI data platform, likely enhancing Appen's human-in-the-loop capabilities amid ongoing demand for quality training data in generative AI and multimodal models.[4] Trends like edge AI, real-time labeling, and ethical data sourcing will shape its influence, potentially evolving toward more automated hybrids while maintaining human oversight for complex tasks. This positions it as a foundational player in sustaining AI's growth, tying back to its origins in bridging human intuition with machine scale.
Key people at CrowdFlower.
