Rerun

Product Introduction

1. Definition: Rerun is a no-code/low-code AI agent orchestration platform. Technically, it is a cloud-based workflow automation tool that leverages large language models (LLMs) to create autonomous, task-specific AI agents that can interact with external tools and APIs.
2. Core Value Proposition: Rerun exists to democratize the creation of transparent, always-on AI agents for business process automation. Its primary value is making AI agents accessible without coding, while providing full visibility and control, thereby solving the "black box" problem common in AI automation.

Main Features

1. No-Code Agent Builder: Users describe tasks in natural language, and Rerun's platform configures the underlying agent logic. It connects to user-provided tools (like Gmail, Slack, CRM systems) via API integrations, abstracting away the need for manual scripting or prompt engineering.
2. Persistent, Autonomous Operation: Agents are designed for 24/7 execution. They operate on event-based triggers (e.g., new email, time of day, database change) or scheduled intervals, eliminating the need for a manual "run" command. This is powered by a serverless backend architecture on private, per-workspace servers.
3. Human-in-the-Loop (HITL) Approvals: Before executing sensitive actions (like sending a refund or a drafted email), agents automatically pause and request human approval via the Rerun dashboard or integrated Slack notifications. This creates a controlled automation loop with mandatory oversight for critical decisions.
4. Real-Time Observability Dashboard: The platform provides live monitoring of all agent activity. Users can view real-time logs, execution steps, token usage, handoffs between agents, and performance metrics. This feature is central to Rerun's transparency promise, allowing users to audit every action.

Problems Solved

1. Pain Point: The opacity and complexity of building reliable AI automation. Traditional methods require significant developer resources for API integration, prompt tuning, and building monitoring systems, resulting in fragile, unobservable "black box" agents.
2. Target Audience: Non-technical business operators (Growth Managers, Operations Leads), small business owners, and product teams (like Designers, Researchers, Copywriters) who need to automate repetitive tasks but lack coding expertise or engineering support.
3. Use Cases:
   • Accounts Receivable: Automatically chase unpaid invoices by drafting and sending reminder emails, then pausing for approval before issuing a refund.
   • Lead Qualification: Autonomous agents parse incoming lead emails or form submissions, research the company, and draft personalized follow-ups.
   • Inbox Management: An agent sorts and prioritizes emails daily, drafting responses for routine queries and flagging urgent items.
   • Content Workflow: Handoff automation between agents for research, drafting, design briefs, and publishing schedules.

Unique Advantages

1. Differentiation: Unlike monolithic RPA tools or generic chatbot builders, Rerun specializes in LLM-powered, multi-step agentic workflows with baked-in transparency. Compared to using raw OpenAI API, it removes the need to build orchestration, tooling, and monitoring infrastructure from scratch.
2. Key Innovation: The combination of a no-code interface for defining complex agent behavior with a per-workspace private server and a real-time visual execution trace. This "observability-by-design" architecture and the principled HITL integration are its core technological differentiators.

Frequently Asked Questions (FAQ)

1. What is Rerun AI used for? Rerun AI is used for building autonomous AI agents that automate repetitive business tasks like invoice chasing, lead qualification, email management, and internal workflows, all without requiring any programming knowledge.
2. How does Rerun ensure my data is secure? Rerun provides each workspace with its own private, isolated server. You can use the included AI model or connect your own API key (e.g., from OpenAI or Anthropic), ensuring your data never transits through third-party training datasets.
3. Can I integrate Rerun with my existing tools like Slack and Gmail? Yes, Rerun is built for tool connectivity. You can connect your agents to essential workplace tools like Gmail, Slack, CRM software, and databases via API to read information and perform actions autonomously.
4. What happens when an AI agent needs to make a sensitive decision? Rerun's human-in-the-loop system automatically pauses the agent before any sensitive action (like sending money or a final email). It sends an approval request to you via the app or Slack, and the agent only proceeds after receiving explicit approval.
5. Is Rerun a replacement for human employees? No, Rerun is designed as a productivity augmentation tool. It automates repetitive, time-consuming tasks and handles initial research or drafting, but it requires human oversight for approvals and complex judgment, freeing up teams for higher-value work.