Snaplii Raises $15M to Expand AI Agent Gift Card Payments

The $3.15 left over in Snaplii's Target shopping demonstration has a specific destination: a merchant gift card in the customer's wallet. That small remainder explains a surprisingly consequential part of the Toronto fintech company's move into AI payments. Software can choose a payment instrument, and the person behind the software inherits the balance after the order is finished.

Snaplii has raised $15M, with Hat-Trick Capital among its backers, according to Axios's November 5 report. Founder and CEO Spencer Xu is extending a consumer gift-card and cashback business into payment access for AI agents. The company describes an approach built around money the user has already set aside and payment methods merchants already accept.

The commercial opportunity reaches beyond completing a checkout. Snaplii is asking agent developers and customers to trust a separate service with the permissions, available funds and merchant value surrounding the purchase. Existing retail acceptance gives that service a starting point, while the money remaining after each transaction gives it an ongoing relationship to manage.

What Snaplii's $15M Funding Supports

Snaplii's financing arrives as the company expands its Agent-to-Merchants, or A2M, offering. Shopifreaks' funding coverage describes plans to add merchants and agent platforms and expand the US team. Those are intended uses of the new capital, with the size and economics of the agent business still to be established publicly.

The accessible original funding report identifies Hat-Trick Capital as an investor without specifying a lead investor or a round designation. The reported amount is $15M; its currency denomination is not specified in the accessible announcement. Assigning a series label or converting the amount would add precision the evidence does not supply.

The company brings a consumer business into this expansion. Its official history reports 350,000 registered users and more than 500 partner brands across North America. Those figures describe Snaplii's broader footprint, rather than the number of people authorizing agents to buy on their behalf. A familiar merchant network can help a new product reach checkout, while the customer's willingness to delegate the purchase remains a separate commercial question.

For Spencer Xu, the choice is to make that established acceptance network useful to another kind of buyer. An agent platform gains access to a payment route without having to negotiate a new checkout method with every store. Snaplii gains an opportunity to serve transactions initiated somewhere outside its own shopping app. DevCuration's coverage of firmly's merchant AI commerce funding provides an adjacent view of the work involved in bringing retail businesses into agent-driven shopping.

How Prepaid Access Changes the Agent's Job

Snaplii's A2M documentation describes a user funding a balance and defining the agent's spending permissions. The agent can then obtain merchant-ready value within that boundary. Gift cards and prepaid instruments give the software a way to complete a purchase while keeping the user's underlying card credentials outside the agent's reach.

The company's public developer repository documents permission scopes, revocable access and spending limits fixed when a key is created. It also describes different Canadian and US catalogs and purchasing capabilities. For developers, that makes payment access a service with explicit conditions, rather than a prompt asking a model to behave responsibly with an unrestricted financial credential.

This design changes where a customer makes the decision to trust an agent. The user first allocates money and authority, then allows the software to work within them. The product's usefulness depends on making that allocation understandable: how much can be spent, which actions are allowed and how access can be withdrawn.

There is still an entire order around the payment. Software can choose an unwanted item, encounter unavailable stock or face a purchase that needs human clarification. A prepaid limit can constrain the funds available to that software; the shopping experience also needs a clear record of the decision and a route back to the person responsible for it. Payment controls become one part of a larger customer service obligation.

The Gift Card Leaves a Second Decision Behind

Snaplii's October 8 company-run demonstration records an agent buying a $30 Target gift card to cover a $26.85 order. The company says the remaining $3.15 stays on that card in the user's Snaplii account. Its published disclaimer explicitly says the companies featured in the demonstration do not endorse Snaplii and are not affiliated with it.

That example makes the denomination of a payment instrument commercially relevant. The price of the goods and the amount allocated to a merchant card can be different. An agent deciding how to pay can therefore influence where the customer's remaining purchasing power sits, even after it has successfully found and ordered the product.

For a shopper, a clear account of the remaining value matters as much as a successful confirmation message. For an agent builder, the receipt needs to connect the item, the payment instrument and any balance still available. The relationship continues when the user returns to buy something else and expects the software to understand money already committed to that merchant.

Merchants receive a different proposition from a familiar payment method. Existing gift-card acceptance can reduce the integration work associated with a new shopping interface, which is part of the appeal of Snaplii's approach. The wallet and agent still have to carry the complexity around denominations, available offers and the customer's remaining value. A simpler checkout for the store can leave meaningful product work elsewhere in the transaction.

Expansion Means Making the Wallet Useful Again

Snaplii's consumer roots give the company a reason to care about repeat spending. Its wallet already combines merchant gift cards with rewards, while the agent offering brings another interface into the purchasing process. The expansion described alongside the funding connects those interests: more merchant choices, more software able to use them and more occasions for a customer to return.

The relevant adoption questions become specific quickly. Can a customer understand what an agent is permitted to do? Can the software account for a gift-card balance on the next order? Can the payment record make an exception clear enough to resolve without reconstructing the entire conversation? These are operating implications of the documented model, rather than additional capabilities established by the funding announcement.

For payments and commerce operators, Snaplii offers a concrete example of established retail infrastructure finding another use. The company is carrying a human shopping instrument into a software-driven buying process, with the permissions and residual balances that instrument brings along. As its merchant and agent relationships expand, the next instruction may arrive with part of the purchase already funded on a card the customer owns.