Job VC

We are looking for a Senior Python Engineer to join a real-time AI voice/chat platform project. The product combines conversational AI, live audio streaming, speech recognition, text-to-speech, telephony, SIP/FreeSWITCH integrations, and external tool/action execution for AI agents.
The engineer will work on new product features, refactoring and performance optimization of the existing Python backend, and improvements in real-time voice call processing.
Responsibilities

Build and improve real-time AI voice interactions over WebSocket, SIP, Twilio, and FreeSWITCH-based infrastructure.

Work with streaming audio pipelines: STT, TTS, VAD, audio normalization, encoding/decoding, buffering, and low-latency processing.

Integrate and optimize AI providers and SDKs such as OpenAI, Gemini, Anthropic, Deepgram, ElevenLabs, Azure Speech, and related APIs.

Improve Python async and threaded components, including asyncio, background workers, queues, WebSocket handlers, and CPU/GPU-bound audio processing.

Refactor existing backend modules to improve reliability, readability, testability, and performance.

Design and implement new AI agent features: tool calling, knowledge retrieval, action execution, call routing, DTMF handling, and conversation workflows.

Work with SIP/FreeSWITCH call flows, outbound/inbound calls, call transfers, hangups, status tracking, and media streaming.

Debug production issues related to latency, concurrency, audio quality, memory usage, external API failures, and call state management.

Collaborate on architecture decisions for scalable AI voice infrastructure.

Required Experience

Strong senior-level Python experience in production backend systems.

Deep knowledge of asyncio, multithreading, queues, task orchestration, cancellation, timeouts, and concurrency debugging.

Hands-on experience with real-time audio processing, streaming audio, PCM/?-law, sample rates, audio chunks, silence detection, and latency-sensitive systems.

Experience with AI/LLM SDKs and APIs, including streaming responses, function/tool calling, structured outputs, retries, pricing/usage tracking, and provider abstraction.

Practical experience with STT/TTS/VAD technologies such as Deepgram, Whisper, Azure Speech, ElevenLabs, Silero VAD, or similar.

Experience with FastAPI, WebSockets, Redis, MySQL, S3-compatible storage, and background processing.

Ability to read, refactor, and stabilize an existing complex codebase without over-engineering.

Strong debugging skills across async code, network services, external APIs, and audio pipelines.

Strong Plus

Experience with SIP, FreeSWITCH, RTP/media streaming, SIP gateways, or telecom call routing.

Experience with Twilio Voice Media Streams or similar programmable voice platforms.

Experience optimizing ML/audio workloads with PyTorch, ONNX Runtime, CUDA, or GPU-aware processing.

Experience with RAG/knowledge systems, Qdrant/vector search, BM25, and hybrid retrieval.

Experience designing reliable AI agent systems with tool execution, external integrations, and stateful conversations.

DevOps experience with Docker, Gunicorn/Uvicorn, observability, and production incident debugging.

Ideal CandidateThe ideal candidate is not just a Python backend developer, but someone comfortable at the intersection of AI, audio engineering, telecom, and high-concurrency backend systems. They should be able to ship new features, improve architecture, reduce latency, debug difficult production issues, and bring senior engineering judgment to a rapidly evolving AI voice product.