# Ionic Launch: Opening the economy to AI agents

Justin McCarty, Ionic Commerce

AI Engineer World's Fair 2024

[Talk page](https://ai.engineer/talks/Ju9PeKEKb24-ionic-launch-opening-economy-ai-agents) · [Video](https://www.youtube.com/watch?v=Ju9PeKEKb24)

## Ionic Launch: Opening the economy to AI agents

An agent that can recommend a product still needs the data, transaction path, and economic incentive to buy it on your behalf.

### Ordering an Uber should not require an advertising funnel

[0:18](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=18s)

An agent should be able to order an Uber to the airport, find a Mother's Day gift, or book a hotel for a last-minute weekend trip. Each request asks for an outcome: arrange transportation, choose something suitable, secure a room. Ionic starts with e-commerce as a way to let agents perform those kinds of tasks on a user's behalf.

The obstacle is the infrastructure between a request and its completion. Justin McCarty, Ionic's CEO and co-founder, describes an online economy built over thirty years around ads rather than actions. Advertising helps a person discover something and directs them elsewhere; an agent needs enough information and authority to finish the task. **The useful unit of interaction is the completed action, not the click.**

### From a product ad to usable purchase information

[1:00](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=60s)

The first example is a product data feed. Its record serves an ad: a human sees it, clicks through to another site, perhaps visits somewhere else, researches the product, and eventually returns to purchase. The record begins a browsing journey rather than containing everything needed for a decision. McCarty wants the agent to keep that work in one place.

That requires more precise product information. A shopper's fit preference might be a shirt measuring twenty-one inches from top to bottom, rather than the label “extra large.” A fabric preference might require seven percent modal and a particular cotton percentage, rather than the broad description “cotton blend.” A structured representation of the two specified constraints could look like this:

```json
{
  "product_type": "shirt",
  "requirements": {
    "top_to_bottom_inches": 21,
    "modal_percent": 7
  }
}
```

The cotton percentage would also need an explicit value before it could become a matching constraint. **Detailed preferences only help if product records expose comparable attributes.**

The feeds McCarty describes also lack stock and shipping information. Even a product that matches the shopper's preferences is not yet an actionable choice if the agent cannot determine whether it is available or how it will arrive.

| Information | Coarse or missing input | What the agent needs |
|---|---|---|
| Fit | Extra large | Top-to-bottom measurement |
| Fabric | Cotton blend | Material percentages |
| Availability | Unknown | Current stock information |
| Shipping | Unknown | Shipping information |

The selected slide completes this comparison by highlighting availability and shipping in the expanded product record. The criticism concerns the feeds in this example, not a universal inability of advertising feeds to carry those fields.

![Slide titled “Data for actions, not ads” compares two product data records, with availability and shipping highlighted in the expanded record on the right.](https://ai.engineer/api/talk-essay-media/541d43d9cc02c66a16da3d3d29ef24dd769a21f725009faa0124625691493460)

Product data for actions, not ads: availability and shipping join the basic product fields.

### Enrich merchant feeds and expose them through an API

[1:48](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=108s)

Ionic's proposed data layer begins with merchant relationships. McCarty reports partnerships with hundreds of merchants supplying product feeds. An agentic workflow enriches those feeds with editorial content, product reviews, and other information useful to a purchasing agent. The mechanism is aggregation and enrichment: bring the evidence an agent would otherwise seek across multiple sites into one place.

The next concern is how agents access that information. McCarty lists relational databases, vector stores, and a graphical database among the possible formats, while leaving the access design open to continued development with builders. These formats serve different retrieval needs: structured attributes support explicit constraints, while richer product descriptions and reviews provide material for broader relevance judgments.

At the [2024 AI Engineer World's Fair launch](https://ai.engineer/worldsfair/2024/schedule), McCarty announced that Ionic's API was live and available to attendees. That is the historical availability claim behind the presentation; it does not establish present service availability.

### Keep checkout in the decision layer

[2:36](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=156s)

Once the agent has enough information to choose a product, sending the user into another browsing flow would reintroduce the original friction. Ionic's next step is to put the transaction in the same layer as the decision.

The example is headphones for McCarty's flight later that day. The intended sequence is:

1. The agent helps the user choose headphones.
2. The user confirms that they want to check out.
3. A payment token passes through to complete the transaction.

The slide shows three headphone options, a recommendation for option A, a purchase request, payment-processing text, and a purchase confirmation. It illustrates the intended checkout experience; it does not expose a token value or establish a completed real-world purchase. McCarty's goal is a single user-facing action after the decision, with the transaction handled behind it.

![Three headphone options above a chat recommending option A, a request to purchase, payment-processing text, and “Purchase confirmed.”](https://ai.engineer/api/talk-essay-media/6e9b11be9057a0f0facf9e2698ab051b98936edaa659578287fcd296cd222558)

A headphone recommendation proceeds to a purchase request and confirmation.

### Who pays for the purchasing agent?

[3:10](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=190s)

Better decisions and easier transactions still leave a business-model question: who funds the consumer AI product? McCarty recalls seeing a post about accumulating AI subscriptions during the week he had signed up for his fourth pro subscription that month. His forecast is that $20-per-month AI subscriptions can serve early adopters but cannot sustain mass-market adoption indefinitely.

The proposed alternative is merchant funding. Merchants already have a reason to pay for help finding suitable customers; an agent builder can provide that value by matching a consumer's needs to the right product. The subscription-cost slide places this argument alongside the progression from better decisions to better actions and better monetization.

![Phone mockup displaying Peter Yang’s post listing five AI services at $20 each, beside a slogan about decisions, actions, and monetization.](https://ai.engineer/api/talk-essay-media/f711f21bc5d3660e291bb746c0f3433ea07060f243a2ffbab953008893f7ce5e)

AI subscription costs alongside “Better decisions. Better actions. Better monetization.”

E-commerce is the starting market. Hotel rooms, flights, and broader participation in the economy are the expansion McCarty envisages, rather than capabilities established by the launch example.

### Make the agent's contribution visible to merchants

[3:57](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=237s)

An agent that buys something on Amazon makes an impressive demonstration: people applaud and repost it on Twitter. But McCarty points out that the merchant still pays Amazon, and Amazon receives credit for the transaction. Completing a purchase does not automatically give the agent builder either attribution or compensation.

Ionic's proposed alternative connects the enriched data to the purchase journey. Instead of receiving only a checkout result, a merchant would be able to understand:

- **Data distribution:** What product information went out into the world.
- **Model use:** Which model saw that information and how it used it.
- **Action flow:** How the interaction progressed to a transaction.
- **Customer relationship:** How the relationship continued after checkout.

This visibility is the economic counterpart to the shared decision-and-transaction layer. It would make the agent's role in generating a sale legible to the merchant, rather than leaving the merchant with only a marketplace conversion.

McCarty argues that merchants would prefer to pay builders for that contribution. He also acknowledges that Amazon pays, while claiming it pays less. The presentation supplies no matched commission terms for that comparison; the mechanism being proposed is compensation for attributable customer matching and transactions.

### The launch offer

[4:44](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=284s)

At launch, McCarty reported hundreds of merchant relationships available to builders immediately and a catalog containing millions of enabled SKUs. Those are Ionic's reported launch figures, without a breakdown of catalog coverage or checkout eligibility.

He closes by describing the catalog as AI native and continuing to grow, inviting prospective hires to contact Justin or Owen at Ionic Commerce, and directing builders to `docs.ioniccommerce.com` to get started. The invitation brings the three parts together: use richer product information to make a decision, carry that decision into a transaction, and give merchants a reason to fund the agent that helped.

## Resources

- [Ionic Commerce Python SDK](https://github.com/ioniccommerce/ionic_sdk_python) — Python SDK with product-search examples, authentication instructions and error handling. The README identifies the SDK as beta.
- [Ionic LangChain integration](https://github.com/ioniccommerce/ionic_langchain) — Shopping-tool integration and example agent code targeting the historical LangChain 0.0.x series.
- [Google Merchant Center product data specification](https://support.google.com/merchants/answer/7052112) — Reference for product-feed attributes, including availability, dimensions and shipping information.
- [Amazon Associates commission schedule](https://affiliate-program.amazon.com/help/node/topic/GRXPHT8U84RAYDXZ) — US affiliate commission rates by product category, with qualifying-purchase conditions and special commission programs.

## Slides

### Presenting Sponsor

Microsoft is the presenting sponsor.

[0:05](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=5s) [Microsoft presenting sponsor slide.](https://assets.ai.engineer/sites/aiecode2025/sha256/aa57fe826b994113e53ad89dcab6b424f83493a103206d4d402bf0461a0ed24e.webp)

PRESENTING SPONSOR

Microsoft

### Innovation and Platinum Sponsors

AWS is the innovation sponsor; MongoDB, Google Cloud, and Neo4j are platinum sponsors.

[0:10](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=10s) [AWS innovation sponsor, with MongoDB, Google Cloud, and Neo4j as platinum sponsors.](https://assets.ai.engineer/sites/aiecode2025/sha256/514d0115af40253262cf91445cf8da24360e76e6b3a4f5785a9e880f47a97485.webp)

INNOVATION SPONSOR
aws

PLATINUM SPONSORS
MongoDB®
Google Cloud
neo4j

### Building for Actions, Not Ads

Ionic wants agents to act on people’s behalf, from ordering a ride to finding a gift. Its starting point is e-commerce.

[0:23](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=23s) [Ionic Commerce describes its mission to open the economy to AI agents, starting with e-commerce.](https://assets.ai.engineer/sites/aiecode2025/sha256/9349b652ed918c9fbcfce14f2dda7242dc88a7c4925add7f64c9f91a62a0a366.webp)

IONIC COMMERCE IS ON A MISSION TO OPEN
THE ECONOMY TO AI AGENTS, STARTING WITH E-COMMERCE

BUILDING FOR ACTIONS, NOT ADS

### Data for Actions, Not Ads

Ionic enriches a basic product feed with availability, shipping, reviews and merchant data, giving agents more information in one place.

[1:15](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=75s) [A basic product feed is compared with an enriched feed containing availability, shipping, reviews, and merchant information.](https://assets.ai.engineer/sites/aiecode2025/sha256/0180c5486f798b8a0738c61b33d3c5f766692eb94bca6e16620655bd5ad641d9.webp)

DATA FOR ACTIONS, NOT ADS

{
    "item": {
    "id":"1234"
    "title": "The best shirt you will ever own","description": "Black t-shirt, cotton blend,
    "Link": "https://www.example.com/shirts/best",
    “variant”: “M”
    "price": {
    "value": "29.67",
    "currency": "USD"
    }
    "color": "black"
}

{
    "item": {
    "id":"1234"
    "title": "The best shirt you will ever own",
    "description": "The perfect t-shirt can be an endless search, but with this crew neck, your search is over. Cut for a relaxed fit from a soft cotton-jersey, it's tag-free for added comfort and has quick-dry properties. This tee works solo or as a base layer.,
    "Link": "https://www.example.com/shirts/best",
    “variant”: “M”
    "price": {
    "value": "29.67",
    "currency": "USD"
    }
    "color": "black"
    "availability": "in stock"
    “shipping”: “2-day, free”
    “product review summary”: “Based on the reviews, this T-shirt is highly appreciated for its excellent fit and high-quality material,...”
    “merchant”: “bestshirts.com”
    “merchant rating”: “4.6”
    “merchant conversion”: “21%”
}

### Fully Agentic Self-Healing Catalog

Merchant feeds enter through the Ionic portal, while an internal agent pulls editorial content, reviews and product specs. The diagram connects both database paths to an Ionic API for external agents.

[2:23](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=143s) [Diagram of merchant and autonomous data feeds flowing through databases and the Ionic API to an external agent.](https://assets.ai.engineer/sites/aiecode2025/sha256/e7c09467b02164fdbf7b6e1f0591d48495427832aa7f684fbf0ecc2a6a8cb899.webp)

FULLY AGENTIC SELF-HEALING CATALOG

amazon
REI co-op
Walmart
YETI

IONIC MERCHANT PORTAL
IONIC
ACTIVE PULL

RELATIONAL AND VECTOR DB

EDITORIAL
REVIEWS
ADDED PRODUCT SPECS
INTERNAL AGENT
AUTONOMOUS PULL

GRAPH DB

IONIC
IONIC API
EXTERNAL AGENT

### Ever Growing Catalog with Millions of SKUs

Ionic says its catalog contains millions of SKUs and draws on relationships with hundreds of merchants. The slide shows select partners, including Amazon, eBay, Microsoft and LEGO.

[4:42](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=282s) [A grid of select merchant partner logos beneath a catalog headline.](https://assets.ai.engineer/sites/aiecode2025/sha256/f1df80077ef5e3a7f6c41c9d0020628bf8336d37a3e2f1607fa04d1962a06922.webp)

EVER GROWING CATALOG
WITH MILLIONS OF SKUS

amazon
ebay
BARNES & NOBLE
ANKER
DURACELL
Fanatics
belkin
SHEIN
VAIO
Microsoft
DHgate.com
LEGO
BRAUN
Shopee
Lenovo
ZINUS
GOLF PARTNER
Southwest
OfficeMax
OXO
MEL Science
newegg

SELECT MERCHANT PARTNERS

## Full transcript

[0:00](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=0s) [upbeat music] Um, like you said, I'm Justin, CEO and co-founder of Ionic.

[0:18](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=18s) And at Ionic, we're on a mission to help agents interact with the economy, starting with e-commerce. Why? Uh, well, when we started working with agents, we wanted them to be able to do things for us on our behalf.

[0:30](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=30s) We wanted them to be able to order the Uber to the airport later today. We wanted them to be able to find the perfect gift for mom for Mother's Day, find the best hotel for that last minute weekend trip.

[0:40](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=40s) Um, and when we started working on those types of tasks and those types of problems, we quickly ran into, um, a set of issues. And what we learned is that for the last thirty years, we've built a world for ads and not for actions.

[0:53](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=53s) Agents, we want them to do things on our behalf, and so we need to change that paradigm. So let's give you an example of, you know, kind of what that looks like.

[1:00](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=60s) So, um, code here on the left is the, uh, data feed for a product. It's an ad. That ad is static. It's meant for a human to see it, a human to click on it, go somewhere else, maybe go somewhere else after that, learn some more, come back and make a transaction happen.

[1:16](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=76s) Um, I don't know about you, but I don't need my agent bouncing from site to site to do things for me. I want it to all happen in one place.

[1:22](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=82s) Uh, and so what does that mean? Well, it means in the future, I want my agent to know that I don't wear an extra large shirt. I wear a shirt that's measured twenty-one inches from the top to the bottom.

[1:31](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=91s) That I don't need a cotton blend. I need seven percent modal. I need a certain percentage of cotton, right? And so those are the types of things that we can have our agents do.

[1:39](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=99s) But where we're starting today, these ad feeds on the left, they don't even know if a product is in stock. They don't know the shipping. They're static. They're not dynamic.

[1:48](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=108s) So what are we doing? Um, we're building an ecosystem where we have partnerships with hundreds of merchants. Uh, they give us access to that product feed that they have, and then we are using agentic workflow to enrich that data.

[2:02](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=122s) With the loop here on the bottom, we're taking editorial, product reviews, anything else that we need or that your agent will find interesting and putting that all in one place.

[2:11](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=131s) Importantly, it's in the format that an agent wants to see, whether that's a relational database, a vector store, a graphical database. We'll continue to grow and, and work with everyone here as, as, uh, we continue to progress on, on how that data should be accessed.

[2:27](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=147s) That's all accessed via our API. That API is live today and is usable by anyone in this room. Um,

[2:36](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=156s) but after we have the data in the right place to make a decision, we then-- again, like I said, we don't need to be bouncing around. We want that transaction to happen, that action to happen right in the same layer.

[2:47](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=167s) And so from here you can see this example of me potentially finding headphones for my flight later today. Um, and once we've made that decision, my agent's worked with me on my behalf, we can just say, "Yeah, I want to check out," and the payment token can come through, and we can complete the transaction immediately.

[3:05](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=185s) That's the way it should be. Agents should be able to do all these things in one simple step.

[3:10](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=190s) Um, okay, so we talked about better decisions. We've talked about better actions. Last part, the secret, we'll talk about better monetization. Um, okay, so consumer AI products, I think I saw this tweet on the left, the, the week I had signed up for my fourth pro subscription of the month, and it just really resonated with me that

[3:27](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=207s) we're not going to be able to charge consumers for our AI products a twenty dollar a month subscription fee forever. It'll work for the early adopters, but it's not mass market.

[3:36](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=216s) Um, the secret is that merchants will pay everyone in this room to help them find the right consumer, um, and match them to the best products. We're starting with e-commerce today.

[3:47](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=227s) Again, like I said, eventually that'll be hotel rooms. It'll be flights. It'll be everything. But, um, you know, I'll give you an example. We've seen some cool demos out there where people are sending agents out.

[3:57](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=237s) They're going, they buy something on Amazon, and we all like clap, and we repost it on Twitter and like, yes, that is cool. But for that merchant on the other side of the transaction, they're just paying Amazon, and they're get-- Amazon's taking all the credit.

[4:09](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=249s) And so if you were going through the process of enriching a data feed and we're helping agents understand what consumers want, and we're making better decisions happen and transactions happen, instead of a merchant getting a yes, you checked out from Amazon and that being the end of the day, they know what data was going out into the

[4:26](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=266s) world, what model saw it, how they used it, what the action flow was, the transaction took place, and then they're able to understand from that customer the ongoing relationship.

[4:36](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=276s) And they're willing to pay for that, and they'd rather pay you for that than Amazon. Um, side note, Amazon will also pay you as well, just not as much.

[4:44](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=284s) Um, okay, and so to get that started, we've created relationships with hundreds of merchants. Um, you can, can work with them immediately today. Uh, we've got a catalog that's already enabled with millions of SKUs.

[4:56](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=296s) Like I said, it's AI native, and it keeps growing. Um, we're hiring if you want to work with us. It's Justin or Owen at Ionic Commerce. If you want to get started, check out the docs.

[5:06](https://www.youtube.com/watch?v=Ju9PeKEKb24&t=306s) It's docs.ioniccommerce.com. Thank you. [upbeat music]
