QueryFlow

QueryFlow is a no-code AI workflow automation platform that lets users design and run complex AI pipelines through a drag-and-drop interface and conversational chat. It focuses on connecting documents, web search, and LLM reasoning into a single orchestrated flow.

Features

• Visual Workflow Builder: Drag-and-drop nodes to compose AI tasks such as retrieval, summarization, classification, and web search.
• Document Intelligence: Upload PDFs, text, and other documents for automatic extraction, chunking, embedding, and retrieval.
• Context-Aware Chat: Chat interface that routes queries through the designed workflow for grounded answers.
• Persistent Workflows: Save, load, and version workflows with stored configuration in a database.
• REST API Layer: Programmatic access to workflows, document uploads, and vector retrieval.

Tech Stack

• React.js (frontend UI and drag-and-drop builder)
• FastAPI (backend APIs and orchestration)
• PostgreSQL (relational storage for users, workflows, and metadata)
• ChromaDB (vector store for embeddings and semantic search)
• Gemini / LLM APIs (reasoning and generation)