INTL
Freelancer
전문가
외주
원격 가능
Expert Full AI Infrastructure Engineer
예산
$250~$750 USD
예상 기간
3~6개월
난이도
전문가
기술 스택
Cloud Computing
AWS
GCP
Azure
DevOps
MLOps
Terraform
Kubernetes
Docker
CI/CD
Deep Learning Frameworks
GPU Computing
Prometheus
Grafana
Secret Management
Data Encryption
Performance Optimization
AI Model Development
AI Model Integration
Python
AI 분석 요약
Nexusgrowth AI를 위한 전문가 수준의 AI 인프라를 처음부터 구축하는 프로젝트입니다. GPU 지원 클라우드 기반의 모듈형 인프라스트럭처-애즈-코드(IaC) 개발, AI 모델을 위한 CI/CD 파이프라인 구축, 그리고 상세한 문서화가 주요 요구사항입니다. MLOps, 클라우드 아키텍처, 보안 및 성능 최적화에 대한 깊은 전문성이 필요합니다.
프로젝트 원문 설명
I need an experienced AI Infrastructure Engineer who can take a green-field concept and turn it into a production-ready foundation for Nexusgrowth AI. The role covers the full life-cycle: architecting the overall framework, selecting and configuring the right deep-learning libraries, wiring everything into our existing services, and hardening the stack for speed and security.
What you will actually deliver:
• A modular, cloud-agnostic infrastructure-as-code repository that provisions GPU-ready compute, storage, networking, monitoring and secret management.
• Continuous integration / continuous deployment pipelines that test, containerise and ship models automatically.
• Documentation and run-books that let any team member reproduce, extend and troubleshoot the environment.
My acceptance bar is simple: new models must move from notebook to automated deployment with one command; latency and throughput targets must be met under load; and sensitive data must remain encrypted in transit and at rest. If you have a proven record of designing, building and optimising AI platforms from scratch, I’d like to see a concise portfolio link and a brief outline of the toolchain you’d propose for this engagement.
What you will actually deliver:
• A modular, cloud-agnostic infrastructure-as-code repository that provisions GPU-ready compute, storage, networking, monitoring and secret management.
• Continuous integration / continuous deployment pipelines that test, containerise and ship models automatically.
• Documentation and run-books that let any team member reproduce, extend and troubleshoot the environment.
My acceptance bar is simple: new models must move from notebook to automated deployment with one command; latency and throughput targets must be met under load; and sensitive data must remain encrypted in transit and at rest. If you have a proven record of designing, building and optimising AI platforms from scratch, I’d like to see a concise portfolio link and a brief outline of the toolchain you’d propose for this engagement.
Freelancer에서 원본 확인
원본 보기