AI Engineer (Software)

TL;DR

AI Engineer (Software): Building and optimizing a self-hosted AI inference stack and developing user-facing agents for family coordination tools with an accent on GPU performance, latency optimization, and full-stack integration. Focus on tuning serving stacks like vLLM and TensorRT-LLM, implementing quantization techniques, and shipping proactive AI agents.

Location: Remote (USA)

Salary: $100,000 - $135,000 / year

Company

A technology company building tools to help families manage routines and transitions through brands like Cozi and OurFamilyWizard.

What you will do

• Run and optimize self-hosted inference serving layers (vLLM, SGLang, TensorRT-LLM) on GPU hardware.
• Implement aggressive optimizations including tensor parallelism, quantization (FP8, AWQ, GPTQ), and speculative decoding.
• Develop user-facing AI agents for proactive nudges, smart suggestions, and scheduling.
• Build internal AI infrastructure, including orchestration, memory, guardrails, and evaluation harnesses.
• Improve GPU utilization and build visibility/observability for AI workloads.
• Develop full-stack features, including quick UIs for rapid prototyping.

Requirements

• 5+ years of production software experience, including applied AI/ML work.
• Proven experience optimizing self-hosted LLMs on multi-GPU hardware.
• Strong Python skills and general engineering fundamentals.
• Experience with agent frameworks (Claude Agent SDK, LangGraph) and RAG.
• Proficiency with AWS, Docker, CI/CD, and monitoring tools.
• Must be based in the USA.

Nice to have

• Experience building internal platforms, Slack apps, or MCP.

Culture & Benefits

• 100% medical premium coverage for employees and 99% for family members.
• 401k with up to a 4% immediate vesting match.
• Paid parental leave.
• L&D stipend and flexible PTO (15-20 days based on tenure).
• Additional paid time off for holidays, winter break, and volunteering.

Hiring process

• Recruiter Phone Interview.
• Hiring Manager Zoom Interview.
• Team Interview.