Welcome to the official Zams API documentation — your gateway to embedding the stateful, autonomous AI agents you've built on the Zams platform into your workflows and products.

Getting Started

This is a lean, endpoint-first documentation. Jump straight in and let's get started.

Untitled

What is Zams?

Zams is an enterprise platform to build AI agents that are designed to automate nuanced business operations. The Zams platform offers robust support for model abstraction, agent-state management, tool chaining and dynamic memory—enabling agents to reason over enterprise-grade data, perform multi-step tasks, and drive context-aware outcomes. With built-in observability, control, and scalability, Zams serves as the de-facto infrastructure for building AI agents for enterprise data.

API Overview

With the Zams API, you can seamlessly automate the creation, scaling, and integration of AI agents into your daily workflows. Here is a breakdown of what's possible:

1. Build: Invoke stateful agents with tool use, memory, and goal-directed reasoning.
2. Tune: Route agent behavior through the optimal LLM for the task with runtime model swapping, fine-tuned model support and latency-aware routing.
3. Evaluate: Use agent telemetry, prompt traceability, and scenario-based evaluation tooling to close the loop between development and deployment.
4. Deploy: Add agent endpoints to your own product with secure API access, RBAC, and audit trails baked in.
5. Scale: Zams is optimized for agents that operate across billions of rows, distributed data, and live streaming inputs — no manual scaling required.

Support

For API support, contact your dedicated account manager or email us on [email protected].