QRev is an open-source platform designed to replace Salesforce by reimagining sales operations through AI agents. Unlike traditional CRMs that rely on manual data entry and static workflows, QRev introduces digital agents—coordinated by a central ‘Qai’ superagent—that automate sales tasks like lead qualification, campaign generation, and customer engagement. Built for modern sales teams frustrated by Salesforce’s cost and rigidity, QRev offers a flexible, extensible foundation where AI agents handle repetitive tasks while humans focus on strategy. The platform is still in early development, actively seeking community input to shape its architecture around role-based agent capabilities and seamless integrations.
QRev is engineered for technical teams, sales operations leaders, and developers who want full control over their CRM stack. It’s not a SaaS product but a self-hosted, modular system composed of three core components: a React-based client frontend, a Node.js backend with MongoDB persistence, and an AI server leveraging LangChain for LLM-powered decision making. By combining vector databases (ChromaDB) and relational SQL backends (SQLite/SQLAlchemy), QRev enables semantic search, context-aware agent behavior, and scalable data routing—making it suitable for teams building custom sales automation pipelines without vendor lock-in.
What You Get
- AI-Powered Sales Agents - QRev introduces autonomous digital agents that mimic sales roles (SDRs, BDRs, Account Execs) and operate through a central superagent called Qai. These agents can interpret user roles, permissions, and CRM data to automate outreach, lead scoring, and follow-ups without human intervention.
- Modular Three-Component Architecture - The system is split into a React client (frontend), Node.js server (backend with MongoDB), and an AI server (LangChain-powered). Each component can be developed, deployed, or extended independently, enabling teams to customize workflows without touching the entire stack.
- CSV-Based Campaign Generation - Users can upload a sample People CSV file to generate targeted sales campaigns. The AI server processes this data to create personalized outreach sequences, enabling rapid onboarding of new leads without manual configuration.
- JWT-Based Authentication System - QRev implements secure token-based auth using environment-generated JWT secrets for both access and refresh tokens, ensuring production-grade security for API endpoints and user sessions.
- Integrated Integrations Framework - The platform supports extensible third-party integrations (via server-side plugins), allowing connection to email, calendar, and communication tools—though specific integrations require manual credential configuration.
Common Use Cases
- Building a self-hosted sales automation platform - A mid-sized SaaS company replaces its expensive Salesforce license with QRev to automate lead qualification and follow-up sequences using AI agents, reducing SDR workload by 60% while maintaining personalized communication.
- Creating a customizable CRM for niche industries - A real estate brokerage uses QRev to build a sales pipeline tailored to property listings and buyer demographics, leveraging CSV uploads to auto-generate campaigns based on property attributes and lead history.
- Problem: Salesforce is too expensive and inflexible → Solution: QRev - Teams tired of paying per-user fees and unable to customize workflows in Salesforce use QRev’s open-source model to build a tailored AI-driven CRM that scales with their business without vendor restrictions.
- DevOps teams managing sales infrastructure at scale - Engineering teams deploy QRev across multiple cloud environments, using Dockerized components to run the client, server, and AI modules in isolated containers while managing MongoDB and ChromaDB as persistent services.
Under The Hood
The project is a modular, multi-language AI platform designed to support agent-based automation, LLM integration, and scalable chat and email systems. It emphasizes extensibility and structured AI workflows through a well-defined architecture and tooling ecosystem.
Architecture
This system adopts a modular monolithic design with distinct projects for AI, agents, chat, and storage, promoting clear separation of concerns. Each module encapsulates specific functionality while sharing core dependencies.
- Modular structure with distinct projects for AI, agents, and infrastructure
- Layered approach for data handling, AI processing, and user interfaces
- Use of factory and strategy patterns to support multiple AI frameworks
Tech Stack
Built primarily in Python with extensive integration of AI and data processing libraries, the platform leverages modern development practices and tooling.
- Python as the primary language with strong AI and data processing libraries
- Poetry for dependency management and Docker for containerization
- Integration with LlamaIndex, OpenAI, and ChromaDB for vector storage
- pytest for testing with support for async and mocked environments
Code Quality
The codebase reflects a mixed quality profile with some well-structured tests and error handling, though inconsistencies and technical debt are present.
- Comprehensive test suite covering key AI and agent modules with varied depth
- Extensive use of try/except blocks for runtime error handling
- Inconsistent code style and structure across Python and JavaScript components
- Signs of duplicated logic in tests and over-engineered module structures
What Makes It Unique
This project stands out through its agent-centric architecture and multi-framework LLM abstraction, enabling flexible deployment and real-time interaction.
- Pluggable agent framework that supports multiple AI providers without core logic changes
- WebSocket-based real-time status updates for scalable agent monitoring
- Configuration-driven UI rendering for dynamic data grid and table customization
- Broad test coverage including integration and mock-based scenarios for AI workflows