CellarTracker today announced CellarChat, an innovative AI-powered chatbot feature now available in beta within their mobile app. This virtual sommelier brings personalized wine recommendations directly to users’ fingertips.
What CellarChat Can Do
As you can probably guess, this is a form of chatbot embedded in the mobile app, and built around your cellar data, that can answer questions such as:
- What should I open for dinner with lamb?
- Which of my wines are ready to drink?
- Do I have a white under $40 that pairs with fish?
(Because nothing says romance like letting a bot pick your rosé)
In the email “Weekend Toast” announcing the new feature, CellarChat is described as “fast, personal.” Obviously since it knows your wine cellar intimately these sorts of recommendations and advice should be helpful. Having all your wines and, now, a virtual sommelier in your hand means one less reason to pull up Google to do a search, only to land on a random blog.
As the announcement notes, CellarChat is currently in beta. So, possibly expect a bug or two.
The AI-Driven Future of Wine Apps
Otherwise this is a welcome (and expected) innovation. One that follows in a long line of continual updates from founder Eric Levine and team that, in my view, catapulted CellarTracker to the top of any best-of wine app list.
There’s all a web UI available if you want to access through your desktop or laptop.
With the emergence of SLMs (Small Language Models) we’re seeing a boom in AI expansion to the edge. That means we can expect to see AI in more places and more devices than would otherwise be possible with the much larger codebases and complexity found in LLMs (Large Language Models). Those typically require extensive computing power (huge data centers). Mobile apps are a perfect use case for SLMs, as they must work within tight RAM and CPU constraints.
While CellarTracker hasn’t disclosed the specific AI technology powering CellarChat, it’s likely using an open-source SLM combined with RAG (Retrieval-Augmented Generation). Just a guess. In this implementation, RAG would enable the model to access and search through your personal wine collection data, then generate recommendations based on that specific information — explaining how it can answer questions like “Which of my wines are ready to drink?”
Additional New Feature: Custom Lists
In addition to the new winebot, CellarTracker has also another new feature to the mobile app: Lists.
These are flexible ways to build, say, a wine list for a party. Or perhaps just an easy way to track a list of favorites, or summer wines, or best of France or Napa. Think of it almost like an Amazon Wishlist.
- Create curated wine lists for parties and events
- Track favorites by category (summer wines, best of France, top Napa selections)
- Build wish lists similar to Amazon’s popular feature
These updates reinforce CellarTracker’s position as an essential tool for serious wine collectors and casual enthusiasts alike, combining comprehensive cellar management with cutting-edge AI technology. It remains my go-to wine app.