Hive Raises $85M to Power AI APIs for Content Moderation & Object Detection

The Growing Importance of Content Moderation and Hive's Solution

Content moderation is increasingly vital for social media platforms, facing mounting pressure to improve their approaches. A startup named Hive is addressing this challenge with a suite of data and image models designed to automate detection of objects and text.

Hive's Funding and Valuation

Hive has secured $85 million in funding, bringing the company’s valuation to $2 billion. This investment will fuel the expansion of its AI-powered APIs, utilized for content moderation and various automated tasks.

Automating Work with AI Models

“Our core focus is on developing AI models that can automate tasks previously performed manually,” explains Kevin Guo, Hive’s co-founder and CEO. “While Robotic Process Automation (RPA) is valuable, it doesn’t address unstructured work where human judgment is essential.” Hive’s models aim to bridge this gap, offering what Guo describes as “near human level accuracy.”

Investment Details and Previous Funding

Glynn Capital spearheaded a Series D funding round of $50 million, with participation from General Catalyst, Tomales Bay Capital, Jericho Capital, Bain & Company, and other investors. Hive also confirmed a $35 million Series C round led by Tomales Bay Capital in 2020, including strategic investments from Bain & Company and Visa. The total funding raised now amounts to $121 million.

A Quiet Rise and Notable Clients

Founded in 2017, Hive initially emerged from a project during founder Kevin Guo’s time at Stanford, evolving from a Q&A platform called Kiwi. Since then, the company has quietly gained traction, attracting clients such as Reddit, Yubo, Chatroulette, Omegle, Tango, NBCUniversal, Interpublic Group, Walmart, Visa, and Anheuser-Busch InBev, totaling over 100 customers.

From Autonomous Systems to Content Moderation

Hive initially focused on image identification for autonomous systems. The company’s origins are often visually represented with images of cars navigating the Golden Gate Bridge.

Shifting Focus to Moderation

Currently, a significant portion of Hive’s work centers around content moderation, encompassing images, text, and streamed audio – which is transcribed and then analyzed. The autonomous car modeling remains a backdrop, chosen for its less sensitive nature compared to content moderation imagery.

The Competitive Landscape of Content Moderation

The need for effective online abuse management has spurred the growth of several startups in the content moderation space. Companies like Sentropy, Block Party, L1ght, and Spectrum Labs are developing platforms to combat harassment and moderate content. Large technology companies are also building in-house tools, as demonstrated by Instagram’s recent launch of new DM abuse prevention features.

The Power of Crowdsourced Data

Hive distinguishes itself through its crowdsourced data collection. Over the past several years, the company has amassed a substantial database by engaging approximately 2 million users. These contributors are compensated – in traditional currency or Bitcoin – for identifying abusive content and other flagged items. Bitcoin has become the preferred payment method for many contributors.

APIs for Automated Workflows

This extensive database powers Hive’s APIs, enabling customers to automate their moderation processes or any workflow requiring rapid identification of objects or text.

Expanding Language Support and Global Reach

The current language learning within the system primarily focuses on English, Spanish, and French. Funding will be allocated to broaden language support and global coverage, unlocking new applications for Hive’s technology.

Innovative Advertising Applications

Hive is exploring new applications, including an advertising approach that delivers ads related to recently viewed content. This method prioritizes user privacy, avoiding reliance on personal data or browsing history, and is gaining interest from brands seeking alternative advertising strategies.

Cloud-Based Machine Learning and Future Potential

The potential of Hive’s AI is a key driver of the recent investment. Its cloud-based infrastructure ensures scalability and extensibility.

Investor Perspective

“Cloud computing adoption is growing, but cloud-based machine learning remains relatively untapped,” states Charlie Friedland, principal at Glynn Capital. “We anticipate cloud-hosted machine learning will be a major component of cloud growth, and Hive is positioned as a leader in this space.”

Challenges and Opportunities

Currently, Hive does not publicly disclose major technology companies as clients, potentially due to NDAs. Guo notes that some companies have been hesitant to adopt AI tools due to the need for human oversight. However, growing concerns about the challenges faced by human content moderators may drive increased adoption of AI-powered solutions.

Strategic Partnerships and Future Growth

Hive’s strategic partnerships with companies like Cognizant, Comscore, and Bain (an investor) facilitate connections with larger tech companies seeking to outsource human moderation tasks. The integration of AI is expected to play an increasingly significant role in shaping and enforcing online abuse policies.

Note: This article has been updated to reflect the two separate funding rounds totaling $50 million and $35 million.