Runtime and Alloy

Alloy and Runtime show up in the same conversations because both promise fewer manual reviews in onboarding and fraud. They do different jobs. Alloy turns identity and risk data into a decision. Runtime agents do the work around that decision that still lands on a person, across every system your team touches. Most teams that use Alloy would run both.

What each does

Alloy is an identity and fraud prevention platform for banks, credit unions, and fintechs. It combines a data ecosystem, an orchestration engine, and a set of AI and machine learning tools. Through one integration you reach a large network of data partners (Alloy cited more than 270 in mid-2026) and build no-code workflows for KYC, KYB, AML and watchlist screening, transaction monitoring, fraud, and credit decisioning. In 2026 Alloy added the AI Assistant, an agent embedded in Alloy that summarizes case data and recommends a decision a person can approve, and perpetual KYB in the UK and Europe, which re-runs business checks when ownership or registry data changes. Alloy says more than 900 financial institutions and fintechs use it.

Runtime is the AI agent harness for payment and fintech teams. Anyone on risk, compliance, onboarding, payment ops, finance, or support builds agents from their SOPs. Agents work on their own isolated computers in your cloud and reach tools through APIs, databases, CLIs, MCP servers, and a browser for portals with no API. They start read-only, stop for approval before anything moves money, and every run is recorded end to end.

Put simply: Alloy is where the decision is made. Runtime is where the follow-up work gets done.

Where they overlap

The overlap is real, and it is in the review queue.

• Case review. Alloy's AI Assistant automates manual review steps for cases inside Alloy, including watchlist hits and fraud investigations. If your entire backlog is Alloy cases and every answer is in Alloy's data, the AI Assistant may cover it.
• Workflow logic. Alloy's no-code workflows already route applications, step up checks, and send cases to review. You don't need Runtime to decide which checks run.

Where they stop overlapping is the edge of Alloy's data. The AI Assistant is built on Alloy's context. A manual review often needs more than that: the customer's transaction history in your ledger, a payout held at the processor, an email thread with the applicant, a document in a state portal, or a note from support. That is the work Runtime agents are built for.

How they work together

A typical flow for an onboarding exception or a lifecycle re-review:

1. Alloy decides or refers. An application, login, or transaction runs through your Alloy workflow. Most are approved or denied automatically. Some are sent to manual review, for example a watchlist hit, a KYB mismatch, or a fraud signal on an existing account.
2. A Runtime agent picks up the case. The referral lands in a queue the agent watches, through Alloy's API, a webhook, or a Slack alert.
3. The agent investigates past Alloy. It reads Alloy's evaluation and then checks what Alloy can't see: account history in your ledger, related payouts at the processor, open support tickets, prior cases on the same business or person, and any registry or partner portal it needs through the browser.
4. It drafts the decision with evidence. The agent writes a case summary following your SOP: what triggered the review, what it checked, what it found, and a recommended action.
5. A person approves in Slack or Teams. Approve, decline, request documents, or escalate. Anything that moves money, like releasing a held payout or applying a reserve, always waits for a person.
6. The agent writes back and closes the loop. It updates the case in Alloy, notes the ticket, messages the customer if your SOP says so, and the full run (every query, tool call, approval, and cost) is kept as one record your sponsor bank or examiner can review.

Over time the agent can turn a solved pattern, like a common address mismatch, into a script that runs deterministically as a template, in your repos.

This is also where the stuck-payment story matters. A flagged account is rarely only a risk problem. Support has the ticket, payment ops has the held funds, finance sees the recon break. With Runtime it is one investigation with one record, and Alloy's decision is one input to it.

When you need both

• Manual review volume is growing faster than you can hire analysts
• Reviews need data from outside Alloy: ledger, processor, tickets, documents, portals
• Risk events touch other teams, like payment ops holding funds or support answering the customer
• You want agents in your own cloud, with your choice of models, for PII and card data
• Your sponsor bank wants one record per case, including what happened after the decision

When Alloy alone is enough

• Your backlog is mostly Alloy cases that Alloy's own data can resolve
• The AI Assistant handles the review volume you have today
• Other teams are not ready for agents yet
• You don't need agents working across systems Alloy doesn't connect to

Frequently asked questions

Is Runtime an alternative to Alloy?

No. Alloy is an identity and risk decisioning platform that pulls KYC, KYB, AML, and fraud data into one decision engine. Runtime does not replace that. Runtime agents work the reviews, exceptions, and cross-team investigations around Alloy's decisions, and use Alloy as one of their tools.

Does Alloy have AI agents?

Yes. Alloy launched the Alloy AI Assistant in February 2026. It is embedded in Alloy's platform and recommends decisions on cases that a person can approve, or that teams can configure to accept automatically.

What does Runtime add if we already use Alloy's AI Assistant?

Alloy's AI Assistant works on the data inside Alloy. Runtime agents work across systems Alloy does not see, like your ledger, processor, bank files, support tickets, and partner portals, and they serve teams beyond risk, such as payment ops, finance, and support.

How much does Alloy cost?

Alloy does not publish pricing. It is sold through sales.

Can Runtime agents run in our own cloud next to Alloy?

Yes. Runtime agent computers run in your AWS, GCP, or Azure account, or you self-host the whole platform. Agents reach Alloy through its API with scoped credentials.