Senior Backend Engineer

Yolk is changing how sales teams win.

We built an in-call, real-time AI coaching assistant that helps sales reps tackle objections and close deals while they are live on the callnot via a report afterward.

We just found Go-To-Market fit, our user base is growing 2x week-over-week, and we are ready to scale. We’re looking for our second backend engineer %28team of 5 total%29 to own our real-time pipelines and work directly with our Head of Engineering and CEO.

 What You’ll Own

• The Real-Time Pipeline: Own end-to-end latency over WebSockets—from live audio chunks %28Deepgram, Pipecat%29 to the coaching signals displayed on the screen.
• The Async Evaluation Engine: Build robust, long-running pipelines that turn long-form audio into structured data reliably %28handling retries, status, and idempotency%29.
• The Simulation Engine: Architect the stateful real-time voice backend simulating buyers for sales rep training.
• Observability & Resilience: Bring distributed tracing, SLOs, and alerting to the core of our AI conversations.

 Tech Stack We Use

• Core: Python 3.12+ • FastAPI • SQLAlchemy 2.0 %28Async%29 • PostgreSQL
• Async & Messaging: WebSockets • RabbitMQ + FastStream • Redis
• AI & Audio: OpenAI, Anthropic, Gemini APIs • Deepgram %28live STT%29 • Pipecat • WebRTC
• Infra: AWS • Kubernetes + KEDA • Grafana/Prometheus • OpenTelemetry

 Who You Are:

• 5+ years of Backend Experience with deep Python expertise and production-proven real-time systems.
• Master of AsyncIO: You know the event loop inside out, what blocks it, and how to handle hundreds of concurrent WebSocket/streaming connections under load.
• Distributed Systems Builder: You design queues and workers with retries and idempotency from day one.
• Startup Mindset: You are a “finisher.” You love extreme ownership, shipping fast, and building code that survives contact with the real world. strong Plus:
•  Experience with real-time voice AI, WebRTC, or conversational audio pipelines.