Unknown

Senior AI Integration Contractor Duration: Approximately 3 months, with potential extension Location: Remote | Weekly status updates required Position Summary Seeking a Senior AI Integration Contractor to design and deploy a secure, centralized AI gateway on clients internal infrastructure. The gateway will provide a unified interface for approved AI models, initially AWS Bedrock, while supporting future integration of self-hosted open models. The platform must support secure processing of sensitive clinical data and provide centralized routing, access control, PHI/PII protection, auditing, monitoring, and prompt management. Key Responsibilities • Evaluate and select an enterprise AI gateway such as LiteLLM, Kong AI Gateway, or similar. • Deploy and configure the gateway using Docker/Kubernetes on internal servers. • Integrate and securely route requests to AWS Bedrock using a consistent OpenAI-compatible API. • Configure secure AWS connectivity, IAM, PrivateLink, encryption, region controls, and restricted network egress. • Implement PHI/PII redaction using tools such as Microsoft Presidio with fail-closed protection. • Establish RBAC, audit logging, usage controls, rate limits, and cost monitoring. • Implement observability using tools such as Langfuse, Prometheus, OpenTelemetry, Grafana, and Bedrock monitoring. • Centralize prompt management, versioning, and testing without requiring application code changes. • Publish a documented and versioned gateway API and working mock within the first two weeks. • Prepare deployment documentation, operational runbooks, security procedures, and conduct knowledge transfer at project completion. Required Qualifications • Proven experience implementing enterprise AI gateways or LLM orchestration platforms. • Strong AWS Bedrock experience, including APIs, IAM, security, and PrivateLink. • Advanced Docker and Kubernetes experience. • Experience integrating on-premise systems with cloud AI services. • Strong knowledge of HIPAA/HITECH, PHI/PII protection, and secure data handling. • Familiarity with NIST SP 800-53 Rev. 5 and healthcare data security requirements. • Experience with AI observability, logging, monitoring, and governance. Preferred • Experience with self-hosted AI models and inference platforms such as vLLM, TensorRT-LLM, or SGLang. • Knowledge of LLM routing, caching, and cloud cost optimization. • Terraform or Ansible experience. • Familiarity with vector databases and context management. • Experience with open-model ecosystems such as Gemma, gpt-oss, or Nemotron.