Rasa is an enterprise-focused developer platform that provides CALM (Conversational AI with Language Models) infrastructure and a low-code user interface to help teams build conversational AI assistants. The company began in 2016 as an open-source platform for chatbots, voice apps and other conversational services and says it has been downloaded more than 50 million times. Rasa emphasizes using LLMs to understand language while keeping enterprises in control of conversation logic to reduce risks like hallucinations. Its technology has been adopted by large clients in financial services and telecom, including American Express, Deutsche Telekom, two of the world’s three top banks and two of the largest U.S. banks. Leadership includes CEO Melissa Gordon and co-founder/CTO Alan Nichol; the company pivoted toward enterprise customers after launching as open source. Rasa reports it roughly doubled annual recurring revenue (ARR) in 2023 versus the prior year, which the company cites as evidence of traction. Rasa offers an open-source platform and SDK that developers use to build chatbots, voice apps, and other conversational AI services. Its SDK downloads grew six-fold in the past 14 months, and annual recurring revenue has tripled in the last 12 months. Customers include Adobe, Deutsche Telekom, BMW, Airbus, ENGIE, HCA Healthcare, and Orange. The product emphasizes customizability and data privacy compared with big cloud providers, targeting developer-friendly, meaningful applications rather than drag-and-drop tools for non-technical users. Rasa plans to use new funding to continue investing in its technology, hire more talent, and expand the business. The company was founded in Berlin and is now headquartered in San Francisco and is currently more focused on growth than revenues. Rasa offers an open-source stack that lets third parties design, train and deploy conversational AI (text and voice) chatbots. Thousands of developers use the free open-source tools, while a paid enterprise edition provides customer support, testing and training tools, and production container deployment priced by organization size and usage. Its platform runs on a customer’s own training data and can be hosted wherever the customer chooses, a selling point for organizations prioritizing data control and privacy. Notable customers include Adobe (which built an Adobe Stock AI assistant on Rasa), Parallon, TalkSpace, Zurich, Allianz, Telekom and UBS. The company plans to expand the platform to cover images and is moving its headquarters from Berlin to San Francisco while hiring in sales, marketing and business development. Rasa recently raised a $13M Series A to accelerate product development and go-to-market expansion. Rasa develops an open-source NLU and bot platform that lets customers train, tweak and run custom conversational AI models rather than using out-of-the-box services. The company positions its stack as production-ready software that brings recent machine-learning research into usable libraries for developers. Customers named in the article include UBS, BMW, Swiss insurer Helvetia, and women’s health bot startup Tia. Rasa reported its Rasa Stack surpassed 100,000 downloads, up from 30,000 in September 2017. As a result of new funding, Rasa plans to open offices in New York and hire a community engineer plus additional full-time AI researchers. The company is based in Berlin and had 15 employees at the time of the article.

Total raised

$70M

Funding rounds

4

Latest round

Series C

Latest activity

Feb 2024