Runtime
Runtime provides an AI agent harness for payment companies, enabling teams to turn standard operating procedures into agents that work payment, risk, compliance, and support queues. The platform runs in the customer's own cloud, keeps humans in the loop for money movement, and routes each run to the smallest model that passes, reducing costs over time.
- AI Agents
- Artificial Intelligence
- Financial Technology
- Developer Tools
- Software Only
- Regulatory & Compliance Technology
Funding
Raised to date
$500KRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Founder details are not available yet.
Product
Problem
Payment operations teams face growing queues of transaction-monitoring alerts, ACH returns, chargeback evidence, and support escalations that require manual investigation across multiple systems. Existing AI coding agents lack persistent infrastructure, secure credential handling, and human approval gates needed for production-grade payment work, forcing teams to choose between slow manual processes or risky automation.
Solution
Runtime provides an AI agent harness that turns standard operating procedures into agents that work payment, risk, compliance, and support queues. Agents are triggered from Slack, Teams, email, or schedules, and automatically provision isolated sandboxes with the necessary tools and credentials. The platform keeps humans in the loop for sensitive actions like money movement, and every run is recorded end-to-end for audit. Runtime supports leading agent frameworks, runs in the customer's own cloud, and includes forward-deployed AI engineers to build the first agents with the team.
Target Audience
Payment companies on both sides of the transaction — issuers, program managers, acquirers, processors, gateways, ISOs, and payfacs — whose teams handle underwriting, onboarding, payment ops, disputes, risk, compliance, and support. The platform also serves fast-growing technical companies using AI coding agents for production-grade development work.
Features
- Persistent runtimes that maintain state across sessions, allowing agents to build on previous work like a real developer
- Automatic provisioning of repos, tools, and credentials when work starts, eliminating manual setup
- Multi-channel triggers from Slack, Teams, email, Zendesk, GitHub, and scheduled jobs
- Real-time visibility into every query, file, and action with human approval gates for sensitive operations
- SOC 2 compliant with encryption at rest and in transit, plus configurable data retention and no training on customer data
- Support for multiple agent frameworks including Claude Code, Codex, Cursor, and Gemini CLI
- Self-hosting option that runs the entire control plane inside the customer's VPC with bring-your-own sandbox support for Firecracker, gVisor, or custom execution layers
- Provider-aware routing that selects the smallest model that passes for each run, reducing token spend over time
- Credential proxy keeps real keys outside the sandbox, so a compromised agent has no key to steal
- Egress allowlists block all outbound traffic except approved hosts and IP ranges
- RBAC for both humans and agents with scoped API keys, plus hard and soft budget limits per session or user
- Deterministic bash scripts enforce rules with the same verdict every time, rather than relying on prompts