Quelm

Distributed Workflow Platform

Open Dashboard
AI-Native Workflow Orchestration

Build and orchestrate distributed AI workflows

Quelm is a distributed workflow platform where AI agents execute visually composed pipelines in real time with retries, orchestration, dependency resolution, and live observability built in.

Agent Nodes

Drag onto canvas

LLM Agent

Prompt + reasoning

HTTP Agent

REST API requests

Transform Agent

Shape workflow data

Webhook Agent

Trigger external apps

Workflow Running

3 active nodes • 1 completed

Research Company

LLM Agent

Analyze company websites and generate structured business insights.

gpt-4.1critical

Summarize Output

Transform Agent

Convert raw findings into concise markdown summaries and structured output.

parallel taskretry enabled

Notify Slack

HTTP Agent

Send execution summaries to the engineering Slack channel.

POST webhooknon-critical

Node Configuration

Configure selected node

Prompt Template

Research the company website and summarize: products, funding, team size, and market positioning.

Model

GPT-4.1Active

Temperature

0.2Deterministic

Realtime

Workflow orchestration

Distributed

Worker architecture

Fault-Tolerant

Execution pipeline

Extensible

Agent system

Platform Features

Designed for modern AI orchestration

Quelm combines distributed systems architecture with AI-native workflow tooling to provide a resilient execution platform for autonomous pipelines.

Visual Workflow Builder

Compose distributed AI workflows visually with drag-and-drop nodes and dependency graphs.

Modular AI Agents

Plug in LLM, HTTP, Transform, and custom agents without touching orchestration logic.

Live Run Monitoring

Watch every workflow execute in real time with live node status updates and streaming events.

Distributed Workers

Scale horizontally with independent worker processes powered by BullMQ and Redis.

Automatic Retries

Built-in exponential backoff, dead-letter queues, and fault-tolerant execution.

Dependency Resolution

Parallel execution, sequential chains, and fan-out flows handled automatically.

System Architecture

Distributed by design

Under the hood, Quelm orchestrates distributed task execution through Redis queues, BullMQ workers, event-driven orchestration, PostgreSQL persistence, and real-time observability streams.

BullMQ queues with exponential backoff

Event-driven downstream task dispatch

Persistent workflow execution history

Worker heartbeats and live health monitoring

Workflow Builder

Visual orchestration layer

Redis + BullMQ

Distributed task queues

Workers

Parallel execution agents

PostgreSQL

Execution persistence

Build resilient AI pipelines

Orchestrate distributed workflows, autonomous agents, and realtime execution pipelines with Quelm.