SPS CommerceRemote
Senior AI Software Engineer
United States of America
Full Time
1 hour ago
$87,000 - $130,000 USD
H1B Sponsor Likely
Key skills
JavaScriptTypeScriptPythonAILLMClaudeRAGReactAWSKubernetesK8sS3BedrockGitVersion ControlRisk ManagementCollaboration
About this role
SPS Commerce is a leading provider of cloud-based supply chain management solutions, serving a global network of retail trading partners. The Senior AI Software Engineer will architect AI systems and pipelines that enhance the MAX Chat, Monitor, and Connect platforms, ensuring seamless collaboration across the retail ecosystem.
Responsibilities:
- Design multi-step AI agents that don't just "talk" but take action, integrating external tools and APIs to resolve complex user requests end-to-end
- Write and maintain structured system prompts that keep model behavior predictable and within expected boundaries
- Develop and tune retrieval pipelines using vector search, reranking, and contextual compression to keep responses accurate and grounded
- Setting up observability and evaluation infrastructure to monitor for bias, latency spikes, and cost-per-request
- Identify and mitigate risks like prompt injection and context manipulation as part of normal engineering practice — not an afterthought
- Stay current on LLM techniques and actively contribute to the team's shared knowledge
Requirements:
- A Bachelor's degree plus 5 years of software development experience; a Master's Degree plus 3 years of experience; a PhD without prior experience; or equivalent education and work experience
- Proficiency in React and at least one backend language (preferably Python or JavaScript/TypeScript). Must be language-agnostic and willing to learn new stacks
- Experience with Langsmith or similar AI observability/troubleshooting tools (e.g., AWS observability)
- Experience with RAG (Retrieval-Augmented Generation) pipelines, vector databases, and building/scaling AI Agents
- Familiarity with MCP (Model Context Protocol), Claude Code, and AI-assisted development tools (Co-pilot)
- Hands-on experience with Kubernetes (K8s) and AWS (S3, Bedrock)
- Strong understanding of Containerization
- Understanding of LLM-specific vulnerabilities, such as prompt injection and context-risk management
- Version Control: Git