Why write hundreds of intents by hand, when you can leverage tools and services from Giant Otter that efficiently learn intents, flows, and slots from your existing data.
For example, in a small sample of 100 customer support calls focused on order issues (checking status, modifying, canceling, etc.), we found 86 different paths through the conversation and 116 unique customer intents expressed in nearly 1,500 different ways. How long would it take your designers to replicate that from scratch?
Give up on complex decision trees
Bots based on decision trees or flow charts only work for very simple use cases. As scope increases, the complexity of the tree structure grows exponentially, making it incredibly difficult to edit and maintain.
With Giant Otter, the effort required to train the bot scales linearly with new use cases. And we give you simple tools to make global edits to the bot, cutting maintenance effort dramatically.
How it works
Upload or Generate Examples
Import 50-100 example conversations. If the data you need doesn't exist yet, we can help you create it
Discover the Conversation Model
Don't create complex graphs, flowcharts, or decision trees from scratch. Our toolset discovers the intents and flows in your data, turning guesswork into an efficient, repeatable process. We can learn conversations with hundreds of intents and thousands of flows
Shape the model into Your Bot
You've learned a model of the conversation, now edit the dialog responses to capture the voice and the scope of the bot you envision
Deploy your Bot
Export conversations to major bot platforms like IBM Watson, Amazon Lex, Microsoft LUIS, etc. – or, deploy with Giant Otter's own bot API
Optimize your Bot
Run tests, score the results, and refine your bot