Skip to main content

​
Welcome to Bedrock

Bedrock is an API platform for building autonomous AI agents that can execute tasks, use tools, and maintain persistent memory across sessions.

​
What is Bedrock?

Bedrock handles the hard parts of building AI agents:
  • Persistent Memory: Agents remember conversations across sessions with automatic summarization
  • Tool Execution: Connect agents to SMS, email, calendars, or your own APIs via webhooks
  • Autonomous Runtime: Agents run independently, waking and sleeping on schedules
  • Full Observability: Every LLM call, tool invocation, and decision is traced

​
Core Architecture

┌─────────────────────────────────────────────────────────┐
│                      Organization                        │
│  ┌───────────────────────────────────────────────────┐  │
│  │                    Product                         │  │
│  │  • LLM Provider Keys (OpenAI, Anthropic)          │  │
│  │  • System Prompt                                   │  │
│  │  • Adapters (tool integrations)                   │  │
│  │                                                    │  │
│  │  ┌─────────────┐  ┌─────────────┐                 │  │
│  │  │   Agent 1   │  │   Agent 2   │  ...            │  │
│  │  │  • Memory   │  │  • Memory   │                 │  │
│  │  │  • State    │  │  • State    │                 │  │
│  │  │  • Tools    │  │  • Tools    │                 │  │
│  │  └─────────────┘  └─────────────┘                 │  │
│  └───────────────────────────────────────────────────┘  │
└─────────────────────────────────────────────────────────┘

​
Key Concepts

Products

Top-level containers holding LLM keys, system prompts, and adapter configurations.

Agents

Autonomous AI entities with their own memory, state, and tool access.

Adapters

Integration modules that provide tools to agents (SMS, email, custom APIs).

Tools

Callable functions agents can invoke—either internal handlers or webhooks.

Memory

Hierarchical summarization system that maintains context across sessions.

Tracing

Full observability into every LLM call, tool execution, and decision.

​
Built-in Adapters

Bedrock includes these default integrations:
AdapterDescription
ContactsStore and manage contact information
SMSSend and receive text messages via Twilio
EmailSend and receive emails via AgentMail
TasksTask management with due dates and status
ProjectsHierarchical project tracking
DocumentsStore text documents for agent reference

​
Custom Integrations

Build your own adapters to connect agents to any system:
  1. Create an adapter with a config schema
  2. Add webhook tools that call your API
  3. Verify requests using the tool call secret
  4. Return data agents can act on
See Tools for webhook implementation details.

​
How Agents Run

1. Wake agent (API call or scheduled)
        ↓
2. Load memory context
        ↓
3. Build prompt with tools
        ↓
4. Call LLM (OpenAI/Anthropic)
        ↓
5. Execute tool calls
        ↓
6. Log to memory
        ↓
7. Repeat until agent sleeps
Agents are autonomous—they decide when to sleep and what actions to take based on their tools and memory.

​
Use Your Own LLM Keys

Bedrock uses your LLM provider accounts:
  • You control costs directly with OpenAI/Anthropic
  • Use any model available on your account
  • Prompt caching with Claude reduces costs significantly

Quickstart

Create your first agent in 5 minutes

Authentication

Set up API keys for your product

API Reference

Full endpoint documentation

Building Webhooks

Connect agents to your APIs