Member of Technical Staff (Software Engineer, Multimodal)

The role

You will design and scale distributed systems that carry live voice sessions in production. You will build backend session-worker architecture, orchestration layers, and streaming protocols for multimodal AI experiences.

What you will do

• Design, build, and scale the backend session-worker architecture for realtime voice.
• Own distributed-systems problems including session lifecycle, crash recovery, and multi-region deployment.
• Build provider-agnostic streaming protocols from the backend to the Rust SDK.
• Drive new products and initiatives in voice and multimodal AI from definition to launch.
• Build the orchestration layer for live voice models to delegate work to tools and agents.
• Partner with SDK, client, infrastructure, and model teams.

Required qualifications

• 4+ years of professional software engineering experience building backend or distributed systems.
• Strong experience in Rust, Python, or Go.
• Experience designing and operating production distributed systems including streaming RPC, stateful services, message-driven architectures, and failure recovery.
• Solid understanding of cloud infrastructure such as AWS or equivalent.
• Strong product judgment to translate user problems into technical solutions.
• Genuine interest and adoption of AI products and willingness to learn quickly.

Nice to have

• Experience with realtime media systems: voice, audio streaming, WebRTC, or low-latency transport.
• Full-stack range — comfort building end-to-end, from backend services through client SDKs or web frontends.
• Experience integrating LLMs, speech models, or computer vision into production systems.
• Experience with agent frameworks, tool-calling architectures, or sandboxed execution environments.
• Time spent at a fast-growing startup or on a high-ownership engineering team.

About the company

Perplexity's mission is to power curiosity. The company has scaled into a new paradigm of using AI to transform knowledge into action.