Wit.ai Integration & Workflow Automation
Run Wit.ai on autopilot. Keep the veto.
31 actions
NLP model changes ship before you realize an intent got deleted or a keyword synonym went live across every app. Rills proposes each change; you approve before it touches your models.
Interactive. No signup. 14 days free · approvals always free.
Most automation fires first, asks later. Rills shows you the change before it ships.
Every consequential communication action from Wit.ai arrives on your phone first. Approve in seconds. Decline without explaining yourself. Workflows wait, paused at zero cost, until you decide.
Queue 3
Delete 3 duplicate intents flagged across 2 apps?
Intents match existing ones with >92% utterance overlap
Last modified 47 days ago, no recent training activity
Deletion is irreversible; no backup utterances detected
Free to wait. Free to think.
Approvals and logic don't cost a credit. Pause a workflow for three hours or three weeks. The price is the same: zero. You only pay when something real happens: an AI call, an outbound action.
Approve from your phone in five seconds.
Swipe right when you're sure. Decline when you're not. Between meetings, mid-coffee, on the train. No dashboard to babysit, no inbox triage, and no surprise send you find out about after it's already out the door.
Routine cases graduate themselves.
Every approval feeds a confidence score for that exact workflow shape. The obvious cases (the ones you've green-lit fifty times) start running on their own. The judgment calls still come to you.
About Wit.ai automation
The model you spent weeks tuning can break the moment an entity keyword gets added in the wrong context or a synonym fires that conflicts with another intent — and you won't know until users start getting wrong answers.
When Wit.ai runs unsupervised
A single unreviewed change to your NLP model can send every downstream response off-target before anyone notices. The blast radius is quiet: no error, just wrong behavior.
- Adding an entity keyword to the wrong entity ships a mismatch that affects every utterance hitting that path.
- Deleting an intent removes a trained behavior permanently, with no undo queued for review.
- A new keyword synonym goes live across all apps the moment it's added, not just the one you intended.
- Deleting utterances removes training data that can't be reconstructed from memory alone.
- Updating a Wit.ai app's settings posts changes globally, including to production-facing configurations.
What Rills does inside Wit.ai
With Wit.ai automation connected, Rills queues proposed changes — adding a value to a trait, creating a new intent, or deleting a keyword synonym — and holds them until you approve. Nothing updates your model until you say so.
The intent still gets created; you just see exactly what it contains before it goes out.
Why Wit.ai has no triggers and how Rills fills the gap
Wit.ai doesn't emit events when your models or entities change, so there's no native way to kick off a review flow from inside the platform. Rills fills that gap by letting you schedule checks and chain actions from upstream signals.
- List Utterances on a schedule to catch model drift or training-data gaps before they compound.
- Get Entity Details and Get Intent Details on a recurring basis to surface any configuration that looks inconsistent with your last approved state.
- List Wit.ai Apps after a deployment step in another tool to confirm only the expected apps were touched before approving further changes.
- List Traits periodically to verify no new values were added outside a sanctioned workflow, especially across shared communication models.
What Rills can do in Wit.ai
6 of 31 actions across reads, writes, and updates.
- 01
Create Wit.ai Entity
Define custom entity types in your Wit.ai application to improve how your chatbot or voice assistant understands and extracts specific information from user messages.
- 02
Create Wit.ai Intent
Defines a new intent in Wit.ai to teach your AI assistant what user messages mean and how to respond appropriately. This enables your chatbot to understand and categorize different types of customer requests with greater accuracy.
- 03
Get Intents
Retrieve all intents you've defined in your Wit.ai app to review, manage, or integrate your natural language processing setup. This helps you understand what user intentions your AI model can currently recognize.
- 04
Wit.ai Get Message
Analyzes customer messages to automatically extract what they're asking for, who or what they're referring to, and important characteristics of their request. This helps you understand customer intent without manual interpretation.
- 05
Create Wit.ai Training Utterances
Add training examples with annotations to improve your Wit.ai natural language model's ability to understand customer intent and extract meaningful information from conversations.
- 06
List Utterances
Retrieve all training utterances from your Wit.ai app to view and analyze the examples used to train your AI model. This helps you understand what data your app has learned from and identify gaps in your training dataset.
Common questions about Wit.ai automation
How do I automate Wit.ai entity creation without breaking my NLP model?
Rills lets you automate Wit.ai entity creation by having AI propose each create, update, or delete action before it touches your model. You review the proposed change and approve it. Nothing gets written to your Wit.ai app until you say so, which means a bad entity definition never ships silently.
Can I get mobile approvals before Wit.ai intents are deleted?
Yes. When Rills proposes a delete intent action, it queues the decision in your mobile approval list. You see exactly what will be removed before it goes out. Approve or reject from your phone, wherever you are, and the action only fires after you confirm.
How do I manage Wit.ai app updates across multiple apps without losing track?
Rills can list all your Wit.ai apps and queue proposed updates for each one. Instead of applying changes directly, it holds them for your review. You decide which updates ship and which ones wait, so a wrong configuration never gets pushed to the wrong app.
Can I automate adding Wit.ai keyword synonyms without AI making unsupervised changes?
Rills automates the proposal of synonym additions but does not add anything to your Wit.ai entities without your approval. Each suggested synonym sits in your queue until you swipe to approve. The AI learns which routine additions you consistently approve and can flag those for faster review over time.
How do I review Wit.ai utterances before any model training actions run?
Rills can list utterances and surface proposed actions based on what it finds. Any downstream change, like adding a keyword or creating a trait, waits for your call. You see the proposed action with context before it goes anywhere near your live Wit.ai model.
How is Rills different from Zapier for Wit.ai automation?
Zapier runs Wit.ai actions the moment a trigger fires. Rills holds proposed actions in a queue and waits for your approval before anything ships. If you have ever had an automated flow create a bad entity or delete an intent you still needed, you know why that difference matters.
Does Rills support Wit.ai webhooks or does it need event triggers to work?
Wit.ai does not emit events that trigger Rills workflows, so Rills works on the action side: proposing creates, updates, deletes, and lookups against your Wit.ai apps on a schedule or based on other triggers in your stack. No incoming webhook from Wit.ai is required.
How much does Rills cost for Wit.ai automation?
Approvals and workflow logic are free. You only pay when Rills executes a real action, like an AI call or an external API operation against your Wit.ai account. Reviewing, approving, or rejecting proposed Wit.ai changes costs nothing, so you are never charged for decisions you make before something ships.