Chatbot glue that keeps Rasa from eating your weekend
An enterprise team's attempt to make Rasa NLU actually deployable without a PhD.

What it does Articulate wraps Rasa NLU—plus Duckling, spaCy, and TensorFlow—into a platform for building conversational agents. It adds a custom dialogue manager that promises “deep complex dialog” while claiming to keep things simple. The project comes from Samtec’s internal “Smart Platform Group,” a corporate R&D team that open-sourced their tooling.
The interesting bit The README is almost aggressively modest: no benchmarks, no feature list, no architecture diagram. Just “we built this, it’s on Rasa, here’s a laptop mockup.” That either signals quiet confidence or a project still finding its public voice.
Key highlights
- Built on Rasa NLU’s ecosystem (Duckling for parsing, spaCy for NLP, TensorFlow for intent classification)
- Custom dialogue manager pitched as handling “deep complex dialog”
- Apache 2.0 licensed, backed by a real enterprise (Samtec) rather than a startup
- React-based UI implied by repo topics, though not shown in README
- Active Gitter community channel
Caveats
- README is thin on specifics: no install requirements, no feature list, no API docs in the main page
- “Focus on simplicity in use” is claimed but not demonstrated—no screenshots of the actual interface beyond a generic laptop mockup
- 592 stars suggests niche adoption; unclear how it compares to vanilla Rasa or competitors like Botpress
Verdict
Worth a look if you’re already committed to Rasa but want a pre-integrated stack with corporate backing. Skip if you need detailed docs upfront or want to evaluate without digging into the /docs folder.