# Trigger.dev > Open source platform for AI agents and long-running workflows. Apache 2.0 licensed. Self-host or let us run it for you. Trigger.dev is an open source platform for building AI agents and background jobs in TypeScript. Write your usual async code and deploy it without worrying about timeouts, server management, or scaling. The platform handles retries, queuing, monitoring, and elastic scaling automatically. ## Open Source Trigger.dev is fully open source under the Apache 2.0 license: - **View Source**: Full codebase available on GitHub - **Contribute**: Accept community contributions - **Self-Host**: Deploy on your own infrastructure with Docker/Kubernetes - **No Vendor Lock-in**: Own your data and infrastructure GitHub: https://github.com/triggerdotdev/trigger.dev ## Deployment Options ### Managed Cloud Use our hosted service with a free tier. No infrastructure to manage. Pay-as-you-go pricing. ### Self-Hosted Deploy Trigger.dev on your own infrastructure: - Docker Compose for simple deployments - Kubernetes/Helm for production scale - Full control over data and networking - Air-gapped environment support Documentation: https://trigger.dev/docs/open-source-self-hosting ## Core Platform Benefits ### No Timeouts Write simple, reliable code that runs for hours. No need to split code into chunks for serverless timeout avoidance. Your tasks run until completion. ### No Servers to Manage We deploy your tasks and handle scaling for you. Focus on building features, not managing infrastructure. ### Durable Checkpointing Tasks automatically checkpoint state. If a task is interrupted, it resumes exactly where it left off. During waits, tasks are frozen and don't consume compute. ### Built-in Observability Every task run is traced with OpenTelemetry. Get real-time alerts via email, Slack, or webhooks. Monitor performance and debug issues with detailed logs. ### Elastic Scaling Infrastructure automatically scales to handle your workload. From zero to thousands of concurrent tasks without configuration. ## Build Extensions Customize your runtime environment with build extensions. Unlike restricted serverless environments, Trigger.dev gives you full control over your build and runtime. ### Available Extensions - **Python**: Execute Python scripts with automatic package installation via requirements.txt - **FFmpeg**: Video and audio manipulation with FFmpeg binaries - **Puppeteer**: Browser automation and web scraping - **Prisma**: Database ORM with migrations - **apt-get**: Install any system packages (libreoffice, imagemagick, etc.) - **esbuild plugins**: Custom build pipeline modifications - **audioWaveform**: Generate audio visualizations - **Custom extensions**: Build your own for project-specific needs This means you can run Python ML scripts, process videos with FFmpeg, generate PDFs with LibreOffice, or any other system-level operation - all from your TypeScript tasks. Documentation: https://trigger.dev/docs/config/extensions/overview ## AI Agent Capabilities Trigger.dev is optimized for building AI agents and LLM-powered applications: ### Agent Patterns Supported - **Autonomous Agent**: AI operates independently, making decisions and taking actions - **Prompt Chaining**: Sequential LLM calls building complex reasoning chains - **Routing**: Direct requests to specialized models based on task type - **Parallelization**: Run multiple AI operations simultaneously - **Orchestrator**: Central coordinator managing multiple sub-agents - **Evaluator-Optimizer**: Assess and iteratively improve outputs ### AI-Specific Features - Human-in-the-loop with waitpoints for approvals - Stream LLM responses to frontend via Realtime API - Tool calling with context from runs - Structured inputs/outputs with Zod schemas - Long-running agent loops without timeout concerns - MCP Server for AI assistant integration (Claude, Cursor, Windsurf) ### MCP Server Trigger.dev provides an MCP (Model Context Protocol) server that enables AI assistants to interact with your tasks directly. AI coding tools like Claude, Cursor, and Windsurf can trigger tasks, check run status, and read logs. ## Product Features ### AI Agents Drop-in AI agent infrastructure. Build intelligent agents using TypeScript with human-in-the-loop capabilities, tool calling, and structured inputs/outputs. https://trigger.dev/product/ai-agents ### Realtime Stream task progress and LLM responses to your frontend or backend. Display task status, progress, and metadata anywhere in your app. React hooks and server-side APIs included. https://trigger.dev/product/realtime ### Scheduled Tasks Build reliable cron schedules. Create recurring schedules up to a year without function timeouts. Supports multi-tenant dynamic scheduling. https://trigger.dev/product/scheduled-tasks ### Concurrency & Queues Fair queuing with three modes: one at a time, parallel, or per-tenant. Concurrency keys for custom queue logic per user or tier. https://trigger.dev/product/concurrency ## Runtime Features ### Waits & Human-in-the-Loop - **wait.for**: Pause execution for durations or until specific dates - **wait.forToken**: Pause until external event (approvals, webhooks, user input) - **wait.forHTTPCallback**: Pause until third-party service completes ### Batch Operations - **batchTrigger**: Trigger many tasks at once - **batchTriggerAndWait**: Trigger and wait for all to complete - Efficient for bulk processing workloads ### Task Configuration - **maxDuration**: Prevent runaway processes - **retry**: Configure retry count and backoff - **queue**: Assign tasks to specific queues - **machine**: Choose CPU/RAM for resource-intensive work - **idempotencyKey**: Prevent duplicate execution ## Infrastructure Features ### Multi-Region Workers Deploy tasks globally for reduced latency. Run tasks closer to your users or data sources. ### Static IPs Connect to enterprise APIs and databases that require IP allowlisting. ### Preview Branches Isolated environments per git branch. Test changes without affecting production. ### Environment Management Multiple environments (DEV, STAGING, PROD) with separate configurations and secrets. ### Versioning Atomic versioning that doesn't affect running tasks. Deploy new versions safely. ## Observability - **Live Tracing**: OpenTelemetry-powered traces for every run - **Logs**: Comprehensive logging with structured data - **Tags**: Attach up to 5 tags per run for organization - **Metadata**: Attach and update metadata as runs progress - **Filters**: Sort and find runs by status, environment, tags, date - **Bulk Actions**: Replay, cancel, or manage multiple runs at once - **Alerts**: Real-time notifications via email, Slack, or webhooks ## Technical Details - **Language**: TypeScript/JavaScript (native SDK) - **Runtime**: Node.js with build extensions for Python, system binaries - **License**: Apache 2.0 - **Deployment**: Managed cloud or self-hosted - **Integrations**: OpenTelemetry, Slack, webhooks, MCP ### SDK Installation ```bash npm install @trigger.dev/sdk ``` ## Use Cases ### AI & LLM Workflows - AI agents and autonomous workflows - Prompt chaining and LLM orchestration - RAG pipelines with document processing - AI content generation at scale - Model evaluation and testing ### Media Processing - Video transcoding with FFmpeg - Audio processing and transcription - Image generation and optimization - PDF generation with LibreOffice/Puppeteer ### Data Processing - Large CSV/Excel file processing - Python data science scripts (Pandas, NumPy) - Batch data transformations - ETL pipelines ### Scheduled Operations - Cron jobs and recurring tasks - Report generation - Database maintenance - Cache warming ### SaaS & Multi-tenant - Per-user background jobs - Tenant-isolated queues - Usage metering and billing - Approval workflows ## Target Audience - **TypeScript/JavaScript developers** building production applications - **AI/ML engineers** deploying agents and LLM workflows - **Python developers** via build extensions - **SaaS companies** needing reliable multi-tenant background processing - **Startups** wanting managed infrastructure without DevOps overhead - **Enterprise teams** requiring self-hosting, observability, and compliance ## Documentation - Getting Started: https://trigger.dev/docs - AI Agents Guide: https://trigger.dev/docs/guides/ai-agents - Build Extensions: https://trigger.dev/docs/config/extensions/overview - Self-Hosting: https://trigger.dev/docs/open-source-self-hosting - API Reference: https://trigger.dev/docs/management/overview - Examples: https://trigger.dev/docs/examples/overview ## Resources - Homepage: https://trigger.dev - GitHub: https://github.com/triggerdotdev/trigger.dev - Pricing: https://trigger.dev/pricing - Blog: https://trigger.dev/blog - Changelog: https://trigger.dev/changelog - Discord: https://trigger.dev/discord