XPENG Motors
Senior Software Engineer, Agentic Platforms
$175K–$296K
Santa Clara, CA +1 · Senior · Full-time · Posted Aug 5, 2026
- Location
- Santa Clara, CA
Mountain View, California, United States - Experience
- 5+ years building and operating production distributed systems.
- Education
- Master’s degree in Computer Science, Software Engineering, or a related field.
About the job
Build and operate XPENG’s internal AI engineering platform, combining reliable distributed services and workflow automation with agentic capabilities that connect agents to engineering systems and safely execute actions.
What you'll do
- Design, build, and operate reliable platform services, infrastructure, and pipelines; own new capabilities from requirements through rollout and operations.
- Automate manual engineering workflows and develop agent lifecycle, integrations, semantic data models, multi-agent workflows, and safe action execution.
- Work with customers on onboarding, requirements, production issues, and adoption; establish practices for agent evaluation, observability, and safety.
What they ask for
- Master’s degree in Computer Science, Software Engineering, or a related field.
- 5+ years building and operating production distributed systems, with strength in system design and software engineering fundamentals.
- Proficiency in Python and TypeScript; knowledge of databases, message queues, caching, and API design.
- Hands-on experience with modern AI coding tools and LLM APIs, including tool use/function calling; experience with agentic systems or task orchestration platforms.
Nice to have
- Familiarity with autonomous driving terminology and model development workflows.
- Experience with semantic data modeling, enterprise or operational system integrations, LLM evaluation/observability, or sandboxing.
Pay
$174,720 - $295,680 USD base salary
Tools
Python, TypeScript, LLM APIs, MCP servers
AI focus
Build an agentic engineering platform, including agent lifecycle and orchestration, tool integrations, semantic grounding, safe execution, and evaluation, observability, and safety guardrails.
Model-written summary · the original posting is authoritative