Platform Overview and Core Architecture
Scarlet and Jo is an AI agent platform focused on workflow automation, data processing, and secure task execution. The system uses a modular architecture where users configure agents to handle structured inputs, apply rules, and return verified outputs. The platform emphasizes transparency in decision paths and auditability of agent actions, which supports compliance and operational monitoring. More details on the platform structure are available on the official Scarlet and Jo documentation page Scarlet and Jo documentation.
The core stack combines rule-based orchestration with model-based reasoning, allowing agents to switch between deterministic paths and probabilistic inference depending on task requirements. This design aims to reduce hallucination risk and improve repeatability in finance, support, and internal operations. Agents can be chained together to form multi-step pipelines, with each step logging inputs, outputs, and confidence indicators.
Token Economics and Usage Model
Scarlet and Jo uses a token-based usage model where compute, storage, and agent execution are metered through internal tokens. Token allocation is tied to workspace plans, with higher tiers providing larger quotas for concurrent agents, longer context windows, and higher throughput. Billing is typically based on token consumption per task, and administrators can set per-agent limits to control costs. Current pricing and token allocation details are published on the Scarlet and Jo pricing page Scarlet and Jo pricing.
The token system supports both pay-as-you-go and committed-use options, with volume discounts for sustained workloads. Token consumption is visible in real-time dashboards, allowing teams to track spend per agent, per workflow, and per integration. The platform also provides usage forecasts based on historical patterns, helping finance teams align budgets with projected agent activity.
Security, Compliance, and Integration
Scarlet and Jo implements role-based access control, encryption at rest and in transit, and audit logging for all agent actions. The platform is designed to meet common enterprise security requirements, including SOC 2 alignment and data residency controls, with specific certifications detailed on the Scarlet and Jo security page Scarlet and Jo security. Administrators can define policy rules that restrict which data sources agents can access and which actions they can perform.
Integration options include REST APIs, webhooks, and pre-built connectors for common business systems such as CRM, ticketing, and database platforms. These connectors allow agents to pull structured data, trigger external workflows, and write results back to source systems with configurable approval gates. The Scarlet and Jo developer portal Scarlet and Jo developer portal provides API reference, SDKs, and sandbox environments for testing integrations before deployment.