May Habib, Cofounder and CEO Author (left) and Wasim Alsikh, Cofounder and CTO Author (right) have raised $100 million in a Series B funding round led by Iconiq Capital.
the author
Companies like Spotify, Uber and Accenture use Author’s generative AI tools to research, create and analyze content.
Generative AI startup Writer has raised $100 million in a Series B funding round, valuing it between $500 million and $750 million, the company announced Monday. Author’s large language models produce content from incident reports and emails to product descriptions and executive summaries, competing with fast-growing unicorns like OpenAI’s ChatGPT Enterprise and Typeface, which launched last month.
But despite the crowded generative AI space, author CEO and cofounder May Habib said Forbes Some enterprise customers are switching from Azure OpenAI to Writer because the quality of the output generated by ChatGPT was not high enough. Her startup’s revenue has grown 10-fold in the past two years and four-fold since the start of this year.
“We’re seeing a lot of people reach out because they’re caught up in proof-of-concept refinements,” she said. “We’ve seen some customers where they couldn’t get the use cases to production because the generations weren’t high enough.”
Author, which was founded in 2020 and featured on Forbes’ AI 50 list earlier this year, has 150 enterprise customers, including Uber, Spotify, Vanguard, Samsung, Accenture and L’Oreal. Iconiq Growth led the round with participation from WndrCo, Balderton Capital, Insight Partners and Aspect Ventures, bringing the startup’s total funding to $126 million.
The author offers a set of 14 models of varying sizes, from 128 million parameters to 43 billion parameters — much smaller than the size of OpenAI’s GPT-4, which has about 1 trillion parameters. The models, called Palmyra, were trained on public data from sources such as web pages, books, Wikipedia, Github, and transcribed video content on YouTube. Public data is filtered to remove copyrighted material, Habib said. Each company gets a different refined version of the model trained on company-specific proprietary data such as financial reports and marketing copy. The model also complies with most privacy and security standards, including HIPAA and Europe’s privacy laws GDPR.
Most recently, in July, the author released PalmyraMed, which is trained on public medical datasets such as PubMedQA that include questions and answers and articles for healthcare applications. Writer’s small fine-tuned models are trained at one-hundredth the speed of other large language models and help them perform specific tasks at a faster rate, Habib said.
“The challenge for industries is to fine-tune LLM on the ocean of data they have and come up with a model that can solve each problem,” she said. “It’s really the last step in tuning, data creation and integration into existing workflows that enterprises need help with.”
Habib said that the main selling point of Author is its integration with tools used by employees (Salesforce and Adobe) and that the app can be embedded into workspaces such as Google Chrome, Figma, Google Docs, Canva, Microsoft Word and Outlook by installing Author’s plugins or An extension that automatically marks content with notifications. “The novelty of chat has worn off,” she said.