HumanSignal

HumanSignal enables data science teams to build AI models with their company DNA. With the emergence of generative AI, it's more important than ever to build highly differentiated models by guiding foundation models with proprietary data and human feedback. Creators of Label Studio, the most popular open source data labeling platform, HumanSignal empowers data scientists to develop high quality datasets and workflows for model training, fine-tuning, and continuous validation. The company, formerly known as Heartex, was founded in 2019 with the mission to advance data-centric AI and put the power in data scientists' hands. Open source software is central to this mission, making it easy for anyone to get started, providing flexibility, and fostering community benefits.

HumanSignal believes that internal teams and domain experts should be responsible for annotating and curating training data, reducing bias through collaboration, and fueled by automation at scale. This approach is based on the founders' real-world experiences across different industries. The company emphasizes that as foundation models fuel new possibilities for AI applications, they also present challenges such as lack of differentiation, mistakes, hallucinations, and bias. Therefore, the need for specialized, high-quality training data to fine-tune and continuously validate generative AI models is more critical than ever. Label Studio, HumanSignal's core offering, has been utilized by over 250,000 users globally, who have collectively annotated more than 200 million pieces of data. The platform supports various data types including images, audio, text, HTML, time-series, and video. HumanSignal offers Label Studio Enterprise as a cloud service with enhanced security (SOC 2 Type II and HIPAA compliant), automation, quality review workflows, and performance reporting, serving leading data science teams at companies like Bombora, Geberit, Outreach, Trivago, and Zendesk. The company has secured $30 million in funding from investors including Redpoint Ventures, Unusual Ventures, Bow Capital, and Swift Ventures.