Senior Backend Engineer (AI Deployments)

We are seeking an autonomous Senior Backend Engineer to build and scale our real-time AI avatar platform.

You will own systems that transform user voice input into GPU-rendered, lip-synced avatar video streamed in real time, covering lipsync inference, conversational orchestration, voice-cloning/TTS, and GCP deployment.

🛠️ What You’ll Do
> Real-Time Streaming: Develop and scale WebSocket/Protobuf services for low-latency audio and lip-synced video streaming.
> AI Orchestration: Build and enhance FastAPI/asyncio services coordinating LLM, TTS/voice-cloning, and lipsync services.
> GPU Inference: Manage PyTorch/CUDA inference pipelines, including GPU allocation, warm pools, batching, and multi-user concurrency.
> CPU Avatar Rendering: Develop a CPU-only avatar rendering path to support lower-cost, scalable sessions where full GPU lipsync is not required.
> API & Integration: Design and maintain REST, WebSocket, and Protocol Buffer APIs and service contracts.
> Cloud & Reliability: Deploy and operate Docker-based services on GCP Cloud Run/GCE and GPU cloud providers, focusing on health checks, scaling, cold starts, reliability, and cost.
> System Design: Collaborate with frontend engineers on API contracts and end-to-end data flow.
> Session & Data: Work with Redis for session/state management and GCS/Vertex AI for storage and RAG-based retrieval.

✅ What You’ll Bring
> 5+ years software engineering experience in backend or distributed systems.
> Strong expertise in Python 3.10+, asyncio, type hints, and FastAPI or equivalent async frameworks.
> Hands-on experience with WebSockets, streaming protocols, concurrency, and low-latency systems.
> Strong knowledge of REST APIs and Protocol Buffers/gRPC-style contracts.
> Experience with microservices, inter-service communication, and session-state management such as Redis.
> Hands-on experience with GCP, including Cloud Run, GCE, Cloud Storage and Artifact Registry.
> Proficiency in Docker, Linux administration, deployment and troubleshooting.
> Understanding of GPU inference, including PyTorch/CUDA, GPU memory, batching, and concurrency.

🌟 Nice to Have
> ML inference serving experience with PyTorch, ONNX Runtime, model warm-loading and batching.
> LLM/GenAI API experience, including Vertex AI, Gemini and RAG.
> Knowledge of audio/video streaming, encoding, buffering and backpressure.
> Experience with GPU cloud providers such as RunPod/Vast.ai and cost/latency optimisation.
> Firebase or equivalent authentication experience.
> CI/CD pipeline experience.

🌟 Why Join Us?
> Work on production-scale AI systems for real-world retail applications.
> Take end-to-end ownership in a highly autonomous environment.
> Join a fast-moving team that values innovation, technical excellence and clean code.