
AI Native Software Engineer
SYNCREON
contract
Posted on:
Location Type: Remote
Location: North Carolina • United States
Visit company websiteExplore more
About the role
- Design and implement AI agents, including: ◦ Retrieval (RAG) ◦ Orchestration workflows ◦ Tool/function invocation ◦ Policy-based routing
- Build evaluation frameworks for accuracy, latency, and reliability
- Implement observability and monitoring for agent lifecycle.
- Integrate with AI providers (e.g., OpenAI, Anthropic, Google Vertex, open-source models)
- Build abstraction layers to support multi-model and multi-provider architectures
- Optimize model usage for performance, cost, and latency
- Develop scalable services using: ◦ Microservices architecture ◦ Containers (Docker, Kubernetes) ◦ Serverless and event-driven patterns
- Implement CI/CD pipelines and infrastructure as code (e.g., Terraform, Helm)
- Ensure production readiness, logging, monitoring, and fault tolerance
- Build and deploy AI-powered applications aligned to business workflows
- Integrate AI systems into existing enterprise platforms and APIs
- Develop backend services and APIs supporting agent workflows
- Define and execute test strategies for AI systems
- Measure system performance (latency, throughput, accuracy, cost)
- Debug and optimize production systems.
Requirements
- 8–10+ years of software engineering experience
- Strong experience with cloud-native systems (APIs, microservices, containers, serverless)
- Experience building and deploying AI/LLM-based systems in production (agents, RAG, orchestration)
- Proficiency in Python, Java, or similar backend languages
- Experience with: ◦ CI/CD pipelines ◦ Infrastructure as code ◦ Monitoring and observability tools
- Hands-on experience with AI platforms (OpenAI, Claude, Vertex AI, or similar)
Applicant Tracking System Keywords
Tip: use these terms in your resume and cover letter to boost ATS matches.
Hard Skills & Tools
PythonJavaAI agentsRAGmicroservices architecturecontainersserverlessCI/CD pipelinesinfrastructure as codemonitoring