In our final article, we demonstrated our new AI features. By constructing on prime of those, we are able to construct any quantity of complexity into our AI chatbots — together with a purchasing cart AI chatbot, as illustrated on this article. If you wish to attempt the AI chatbot we’re constructing on this article you’ll find it beneath.
Video Demonstration
Within the following video, I’m demonstrating the purchasing cart AI chatbot, by going via your entire purchasing expertise utilizing nothing however the AI chatbot to buy. As well as, I’m exhibiting extra options virtually inconceivable to implement utilizing a extra conventional UI-based purchasing cart, comparable to the power to inform the AI chatbot the next and have it intelligently reply again to you.
“I’ve got 1 million to spend, what items can I add without exceeding my budget?”
The above video is barely 7 minutes lengthy. I significantly suggest you watch it to grasp the facility of this characteristic.
How It Works
The entire concept of the chatbot is to can help you use pure language to information you thru your whole purchasing expertise. Under is an instance of learn how to add two objects to your cart. For those who look rigorously on the picture beneath you will notice two inexperienced icons. This suggests the AI chatbot invoked two AI features, one for every product I requested it so as to add for me.
Along with including objects, it additionally has the next features.
- Take away the merchandise from the cart
- Record all merchandise
- Record all objects in my cart
- Checkout and pay for my objects
The checkout course of merely creates a abstract e mail that it sends to you, however this half may simply be built-in with WooCommerce or Shopify to carry out an actual sale.
Implementation
An important a part of our AI chatbot is its system message. Our system message incorporates references to AI features, comparable to follows.
## Including objects to purchasing cart
If the consumer informs you that she or he desires to purchase an merchandise and you understand
the merchandise's SKU, then reply with this EXACT response, solely exchanging the
[sku] and the [quantity].
___
FUNCTION_INVOCATION[/modules/shopping-cart-demo/workflows/add-item.hl]:
{
"sku": "[sku]",
"quantity": "[quantity]"
}
___
If the consumer didn't present a amount then don't ask the consumer for a amount
however use a price of 1 by default. If you do not know the merchandise's SKU, ask the
consumer to offer some extra key phrases for what merchandise she or he desires to purchase such
that you'll find the proper SKU earlier than responding with the above operate
invocation.
The above instructs OpenAI to return a FUNCTION_INVOCATION if the consumer says she or he desires so as to add an merchandise to his purchasing cart. Such operate invocations will probably be executed on the cloudlet and comprise references to AI workflows. Under is the AI workflow for the above operate.
/*
* Provides an merchandise to your purchasing cart
*
* [product_id] is necessary and must be an current product_id from an
* merchandise current in your database, and [quantity] will default to 1 if not
* specified.
*/
.arguments
sku:string
amount:int
session:string
.description:Provides an merchandise to your purchasing cart
.kind:public
// Defaulting [quantity] to 1 if not specified.
validators.default:x:@.arguments
amount:int:1
/*
* Invokes the SQL CRUD Learn slot with the required parameters.
*
* Present [connection-string], [database-type], [database], and [table] to
* inform the motion of what database/desk you wish to execute your SQL
* in the direction of, and add [and] or [or] arguments to filter your returns, in
* addition to [limit] and [offset] to use paging. Use [order]
* and [direction] to type both ascending or descending. Discover, you possibly can
* solely use one in all [or] or [and], and never each.
*/
execute:magic.workflows.actions.execute
title:sql-read-products
filename:/misc/workflows/actions/sql/sql-read.hl
arguments
columns
.:product_id
database:shopping-cart-demo
desk:merchandise
and
sku:x:@.arguments/*/sku
/*
* Invokes the SQL CRUD Create slot with the required parameters.
*
* Present [connection-string], [database-type], [database], and [table]
* to tell the motion of what database/desk you wish to execute your
* SQL in the direction of, and [values] in your precise values to insert.
*/
execute:magic.workflows.actions.execute
title:sql-create-shopping-cart-item
filename:/misc/workflows/actions/sql/sql-create.hl
arguments
database:shopping-cart-demo
desk:objects
values
product_id:x:--/execute/=sql-read-products/*/*/product_id
amount:x:@.arguments/*/amount
session_id:x:@.arguments/*/session
// Returns the results of your final motion.
return-nodes:x:@execute/*
Such AI features will also be declared as coaching snippets, permitting you to have actually hundreds of them. And 98% of the above code was created with out coding utilizing our Hyperlambda Workflow characteristic.
The tip results of the above turns into that if we are saying stuff comparable to “I want to buy the flying car” to our AI chatbot, then OpenAI will return a operate invocation declaration that our machine studying kind will execute. Our operate invocation inserts a brand new file into our “items” desk in our “shopping-cart-demo” database. This permits the AI chatbot to maintain observe of which objects have been added. You possibly can see the database schema beneath.
It is a pretty naive purchasing cart, nevertheless it additionally permits for integration with Shopify, WooCommerce, or some other e-commerce platform you could have. So do not be fooled by its simplicity.
Safety
Except you implement this appropriately, you would in concept immediate engineer the AI chatbot to return operate invocations which can be unsafe. We guard towards this by checking if the operate invocation is said both within the system message or the kind’s coaching information earlier than we permit for a operate to be executed.
This fully eliminates “function injection attacks,” that are related in nature to SQL injection assaults. Nonetheless, you continue to should watch out when including operate declarations to your kind, and solely add features you understand for a reality are secure to permit the consumer to execute.
Sooner or later we would additionally implement help for features that require the consumer to belong to some function, along with some authentication and authorization mechanism, to permit for the consumer to log in to the AI chatbot, change his or her rights with reference to what features are allowed to being executed by the consumer. However presently, we do not have something right here. However it is a high-priority characteristic we’re engaged on so much, so anticipate issues to quickly change right here.
Wrapping Up
AI features is the by far funniest and most rewarding factor I’ve labored on since I began working with OpenAI. In a method, it turns a “boring text-producing LLM” right into a fully-fledged AI assistant. Such AI assistants will be created to do unbelievable issues, particularly as soon as we add authentication and authorization to them. Under is an inventory of issues we may simply do utilizing AI features.
- Order a aircraft ticket to New Zealand
- Verify the climate in Los Angeles tomorrow
- What is the inventory worth for Apple
- Ship Mike an e mail and inform him I will be late for our assembly
- I must name Jane, give me her cellphone quantity
- Create a brand new buyer in my CRM system
- And so on, and so forth, and so forth …
All the above would actually be a “15-minute job” utilizing Magic Cloud. For those who’re considering seeing what we are able to do in your group associated to this, you possibly can contact us beneath.