Who Layest is for
Layest is built for product, engineering, and operations teams that want to deploy AI agents into real workflows. Whether you’re a solo developer prototyping an automation or an enterprise team rolling out agents across departments, Layest scales with you.You don’t need machine learning expertise to build agents on Layest. The platform is designed for teams who understand their domain and want AI to work within it — not the other way around.
Key concepts
Understanding these core concepts will help you navigate the platform and get the most out of every agent you build.Agents
An agent is the central building block in Layest. Each agent has a set of instructions, a role, access to capabilities, and optionally a knowledge base. Agents can respond to triggers, call external services, and hand off tasks to other agents.Workflows
Workflows define how an agent — or a chain of agents — moves through a series of steps to complete a task. You can build linear workflows or branch on conditions, loop over data, and handle errors gracefully.Capabilities
Capabilities extend what an agent can do. The three main capability types are:- Integrations — Connect agents to external systems like Slack, Salesforce, PostgreSQL, or any REST API.
- Triggers — Define when an agent runs: on a schedule, via webhook, on a form submission, or manually.
- Knowledge — Attach documents, embeddings, and memory stores so agents always have relevant context at their fingertips.
AI Assistant
The AI Assistant is a conversational interface built directly into your Layest workspace. Use it to query your data, run ad-hoc agent tasks, or explore your collections — all through natural language.Data & Views
Collections are structured data stores that live inside your workspace. Views let you slice, filter, and display that data in ways that make sense for your team. Agents can read from and write to collections as part of any workflow.Observability
Every agent run is recorded. Observability gives you full trace logs, step-by-step execution history, latency metrics, and error reporting — so you can debug confidently and improve over time.Next steps
Workspace Setup
Create your account, configure your workspace, and invite your team — ready in minutes.
Quickstart
Go from a blank workspace to a running agent workflow in under five minutes.