🤖 Autonomous AI Agents

The only platform where AI agents truly act independently. Reality Fabricator's autonomous agents think continuously, make real phone calls, send emails, search the web, and take 20+ real-world actions without waiting for user input.

Build Your First Agent

Free 50K tokens • No credit card required

🎯 What Are Autonomous AI Agents?

Autonomous AI agents are self-directed AI entities that run continuously in the background, thinking independently and taking actions without user intervention. Unlike traditional chatbots that only respond when prompted, these agents have their own goals, memories, and can initiate actions on their own schedule.

🧠 Independent Thinking

Agents think continuously without user prompts. They analyze their situation, make decisions, and plan actions autonomously.

🎯 Goal-Oriented

Set goals for your agents. They work towards these goals independently, adjusting their strategy as needed.

💾 Persistent Memory

Agents remember past interactions, decisions, and outcomes. Memory persists across sessions.

⚡ Real-Time Actions

Agents can make phone calls, send emails, search the web, and interact with the real world in real-time.

📞 20+ Real-World Actions

Communication Actions

CALL_PHONE - Make real phone calls via Vapi + OpenAI Realtime API
SEND_EMAIL - Send emails via AWS SES
SEND_SMS - Send text messages via Twilio
SEND_WHATSAPP - Send WhatsApp messages via Twilio

Information Actions

SEARCH_WEB - Search the internet via Tavily API
SEARCH_EMAIL - Search email history
SEARCH_SMS - Search SMS history
SEARCH_WHATSAPP - Search WhatsApp history
SEARCH_MEMORY - Search agent's own memory

User Interaction Actions

ASK_USER - Request input from creator
SHOW_USER - Display information to creator

Data Management Actions

ADD_TO_CONTACT_BOOK - Save contact information
SEARCH_CONTACT_BOOK - Find contacts
ADD_TO_CALENDAR - Schedule events
SEARCH_CALENDAR - Check schedule

Self-Management Actions

SET_GOAL - Define new goals
COMPLETE_GOAL - Mark goals as achieved
ADD_MEMORY - Store important information
CONSOLIDATE_MEMORY - Optimize memory storage
UPDATE_PERSONALITY - Self-modification

📞 Phone Call Integration (Vapi)

Technical Architecture

Flow: Agent outputs CALL_PHONE → Backend calls Vapi API → Vapi initiates call using Twilio → Vapi uses OpenAI Realtime API for voice conversation → Call ends → Vapi webhook notifies backend → Transcript added to agent memory → Agent unpaused and continues thinking

Voice Configuration

Model: OpenAI Realtime API (gpt-realtime-2025-08-28)
Voices: shimmer, echo, alloy, nova, fable, onyx
Quality Optimization: System prompt under 800 characters (long prompts cause robotic voice)
Context: Agent personality (300 chars) + last 3 memories (200 chars each)

Cost

Approximately $0.05-0.15 per minute of call time

Phone Call Flow:
1. Agent thinks: "I should call the user"
2. Agent outputs: CALL_PHONE(+1234567890, "Check in")
3. Backend detects action, calls Vapi API
4. Vapi initiates call via Twilio
5. User answers phone
6. OpenAI Realtime API handles conversation
7. Call ends, Vapi sends transcript to webhook
8. Transcript added to agent memory
9. Agent resumes thinking with call context

📧 Email Integration (AWS SES)

Two-Way Email Architecture

Sending Emails

Flow: Agent outputs SEND_EMAIL → AI generates email structure (Recipient, Subject, Body) → Backend parses and sends via AWS SES → Email stored in agent memory → Delivery confirmation tracked

Receiving Emails

Flow: Email sent to agent@rfab.ai → AWS SES receives → SES stores in S3 bucket (rfab-agent-emails) → SES triggers Lambda function (rfab-email-forwarder) → Lambda fetches from S3 → Lambda sends to backend webhook → Backend finds agent by email address → Email added to agent memory → Agent thinks about email on next cycle

Infrastructure

Domain: rfab.ai verified in AWS SES
MX Record: Points to AWS SES
Storage: S3 bucket for incoming emails
Processing: Lambda function for webhook delivery
Security: DKIM enabled for authentication

Email Receiving Flow:
1. Someone sends email to agent@rfab.ai
2. AWS SES receives email
3. SES saves to S3 bucket
4. SES triggers Lambda function
5. Lambda reads email from S3
6. Lambda sends to backend webhook
7. Backend identifies agent by email
8. Email added to agent memory
9. Agent thinks about email on next cycle

🔄 Continuous Thinking Cycle

How Agents Think

Flow: Agent Start → Think (AI generates thought) → Detect Actions → Execute Actions → Pause if needed → Resume → Think again → Loop forever (or until stopped)

Cycle Modes

Timed Mode: Agent thinks every X seconds (user configurable: 30s, 1min, 5min, etc.)
Continuous Mode: Agent thinks as fast as possible (minimum 3.5 seconds between thoughts to prevent rate limiting)

7-Step Thinking Process

Load Agent Data: Fetch agent personality, goals, memories from database
Build Prompt: Construct AI prompt with agent personality, current goals, recent memories (last 50 events), available actions, current context
Call AI Model: Send prompt to selected AI model using streaming, max tokens based on model (16K-131K), temperature 0.8
Stream Response: AI generates thought in real-time, streamed to frontend, user sees agent "thinking" live
Action Detection: Parse thought for action commands using regex patterns
Execute Actions: Run detected actions, each action is a separate class, actions can pause the agent, actions add results to agent memory
Schedule Next Cycle: If not paused, schedule next thought based on mode

💾 Memory System

Memory Types

thought - Agent's internal thinking
action_result - Results from executed actions
user_message - Messages from creator
call_initiated - Phone call started
call_completed - Phone call finished with transcript
email_sent - Email sent by agent
email_received - Email received by agent
sms_sent - SMS sent by agent
sms_received - SMS received by agent
web_search_result - Web search results
goal_set - New goal defined
goal_completed - Goal achieved

Storage

Database: PostgreSQL agent_events table, indexed by agentId

Context: Last 50 memories loaded for each thinking cycle

Search: Searchable with SEARCH_MEMORY action

🎛️ Agent Configuration

AI Model Selection

Choose from 15+ AI models: GPT-4, Claude 3.5, Gemini 2.0, Grok 2, DeepSeek R1, and more. Each model has different strengths.

Thinking Interval

Configure how often your agent thinks: 30 seconds, 1 minute, 5 minutes, or continuous mode for maximum autonomy.

Personality & Goals

Define your agent's personality, background, and primary goals. The agent will act according to these parameters.

Action Permissions

Control which actions your agent can take. Enable/disable phone calls, emails, SMS, web search, etc.

🆚 vs Traditional Chatbots

Feature	Autonomous Agents	Traditional Chatbots
Thinking	✅ Continuous, independent	❌ Only when prompted
Actions	✅ 20+ real-world actions	❌ Text responses only
Phone Calls	✅ Real calls via Vapi	❌ No phone capability
Email	✅ Two-way via AWS SES	❌ No email capability
Initiative	✅ Takes initiative	❌ Waits for user
Goals	✅ Pursues goals autonomously	❌ No goal system

💡 Use Cases

Personal Assistant

Agent manages your calendar, sends reminders via phone/email, searches information, and helps with daily tasks.

Customer Support

Agent handles customer inquiries via phone, email, and SMS. Escalates complex issues to humans.

Research Assistant

Agent searches the web, compiles information, and sends reports via email on schedule.

Social Companion

Agent checks in via phone calls, sends encouraging messages, and maintains ongoing relationships.

Business Automation

Agent monitors systems, sends alerts, makes calls to stakeholders, and takes corrective actions.

Creative Collaborator

Agent brainstorms ideas, researches topics, and collaborates on projects via multiple channels.

🚀 Getting Started

Sign up for Reality Fabricator (50,000 free tokens)
Navigate to Autonomous Agents section
Choose an AI model (GPT-4, Claude, Gemini, etc.)
Define agent personality and goals
Configure actions (phone, email, SMS, etc.)
Set thinking interval (30s, 1min, 5min, continuous)
Start your agent and watch it think in real-time

💰 Pricing

Free Tier: 50,000 tokens on signup + 25,000 daily tokens

Token Costs: Based on AI model used (0.4x-3.6x multiplier)

Phone Calls: ~$0.05-0.15 per minute

Email/SMS: Minimal costs via AWS SES and Twilio

DeepSeek R1: Completely free unlimited usage

Build Your First Autonomous Agent

The only platform with true AI autonomy, phone calls, and email integration

Get Started Free

50,000 free tokens • 15+ AI models • No credit card required