Member of Technical Staff

Responsibilities:
Design, build, and continuously improve the stateful conversational agent runtime powering real-time AI interactions.
Develop and optimize the streaming pipeline (SSE/WebSockets), ensuring smooth response delivery, natural pacing, and clean interruption handling.
Architect and maintain conversation state management systems, ensuring user-visible responses, model memory, and persisted context remain perfectly synchronized.
Design and implement the tool orchestration layer, enabling the agent to call domain-specific tools mid-conversation and reason over results in real time.
Optimize system performance for latency (first-token time), reliability, interruption recovery, and context window efficiency.
Debug and resolve complex production issues related to concurrency, race conditions, and distributed coordination.
Requirements:
3-6 years of strong backend or systems engineering experience (high-ownership startup environments preferred).
Deep understanding of state machines, concurrency models, race conditions, and failure handling.
Hands-on experience building real-time or streaming systems using WebSockets, SSE, or similar protocols.
Experience building and maintaining long-running stateful processes in production environments.
Strong debugging skills for complex, real-time production systems.
Proficiency in backend languages such as TypeScript/Node.js, Go, Elixir, or similar.
Familiarity with LLM APIs and streaming responses is preferred.
Experience with actor-model systems (e. g., Elixir/OTPGenServers, supervisors) is a strong plus.
Prior exposure to conversational AI systems or agent-based architectures is a bonus.

Experience
3-6 yrs