Case Study: Meekan achieves in-house NLU, improved performance and data ownership with Rasa

Meekan switched from the cloud-based API Microsoft LUIS to their own in-house NLU with Rasa

Meekan, a Tel Aviv–based Slack bot for scheduling, needed to move its natural language understanding in-house because Microsoft LUIS offered no domain-specific customization and kept their data outside the company. Lacking the resources to build a LUIS-equivalent from scratch, Meekan chose Rasa to gain control over their NLU and ownership of training data.

Using Rasa’s toolkit and fast migration guide, Meekan imported their LUIS training data and went from idea to production in four weeks with no interruption to end users. Rasa enabled Meekan to tailor algorithms to the scheduling domain, improve NLU performance, own their models and data for competitive advantage, and reduce latency for faster responses at scale.

Meekan

Eyal Yavor

Chief Technology Officer