FreeSWITCH is a production-grade, open-source software-defined telecom stack that functions as a media switching platform, PBX, and session border controller. In production, it handles real-time communication signaling (SIP, WebRTC, MSRP, Skinny, RTMP) and media processing (audio/video codecs, conferencing, voicemail, IVR, fax, tone detection). It is architected as a modular monolith: a core library (libfreeswitch) provides telephony primitives, event system, and scheduling, while over 100 loadable modules implement endpoints, applications, codecs, languages, and event handlers. Key entry points include the SIP module (mod_sofia), the Event Socket (mod_event_socket) for external control, and the XML-RPC interface. Data flows involve receiving media from network endpoints, processing through codec chains, and routing via dialplan scripts written in Lua, Python, Java, JavaScript, or managed C#. Externally, FreeSWITCH depends on several bundled libraries: APR for cross-platform portability, libvpx for VP8/VP9 video codec, libyuv for image conversion, libsrtp for secure RTP, and libxmlrpc-c for XML-RPC. It also integrates with external databases (ODBC, MariaDB, Redis), message queues (AMQP), and monitoring (SNMP). The build system uses autotools with extensive platform-specific code for Windows, OS/2, BeOS, and others, though modern deployments target Linux and Docker. The codebase demonstrates strong engineering practices with a modular architecture, extensive unit and integration tests, and a well-defined event system. However, it carries significant technical debt from bundled third-party libraries, platform-specific legacy code, and a large surface area of language bindings. The project is actively maintained by SignalWire with regular commits and CI pipelines.
Dograh is an open-source, self-hostable voice AI platform for building, deploying, and monitoring voice agents. It provides a visual workflow builder (node-based graph editor) for designing call flows, with deep integration across telephony providers (Twilio, Plivo, Vonage, Telnyx, Cloudonix, Vobiz, Asterisk ARI), STT/TTS/LLM models (OpenAI, Deepgram, Google, Azure, Cartesia, ElevenLabs, etc.), and real-time audio pipelines via the Pipecat library. The backend is a FastAPI monolith with PostgreSQL, Redis, and object storage, complemented by separate worker services (ARQ for background tasks) and a campaign orchestrator for bulk outbound calling. The system supports multi-tenancy, call recording, knowledge base/RAG, QA analysis, embeddable chat widgets, and an MCP server for AI-assisted workflow management. SDKs for Python and TypeScript enable programmatic workflow creation and outbound call triggering. While the core is a full-stack application, it includes evaluation frameworks for STT providers and post-call QA analysis, giving it a hybrid character but dominated by production application concerns.