AI/ML Platform Engineer – On-Prem LLM & DevOps

CACI International Inc; Lisle (IL); Sur place; USD 90 000 - 190 000; Plein temps

Résumé du poste

CACI International Inc. is seeking an experienced AI/ML Platform Engineer in Lisle, IL to support on-prem LLM-enabled technologies and agentic workflows.

You will architect the greenfield implementation, deploy and monitor LLMs on local Kubernetes clusters, and build RAG pipelines integrated with MCP servers. The role requires deep AI infrastructure knowledge, strong DevOps practices, and collaboration across cross-functional teams to iterate improvements and ensure secure, scalable model

Qualifications

• Bachelor's degree in CS/IS/Engineering or equivalent with related experience.
• Four years of related experience or a Master's with two years of related experience.

Responsabilités

• Architect a greenfield LLM-enabled tech stack for agentic workflows.
• Deploy, operate and monitor on-prem LLMs on local Kubernetes clusters.
• Design and maintain RAG pipelines for document retrieval and embedding workflows.
• Develop and maintain MCP servers and their integrations with existing services.
• Collaborate to implement CI/CD pipelines and improve product cycles.

Connaissances

IT fundamentals

Communication

Python/SQL/Java

Data pipelines

Kubernetes/Elasticsearch/Terraform

GitLab CI/CD

RAG/MCP/LoRA

Data warehousing

Teamwork

Formation

Bachelor’s degree in CS/IS/Engineering

Outils

Kubernetes

Kserve

vLLM

LangFuse

Elasticsearch

Terraform

Nexus

GitLab CI/CD

Apache Iceberg

DBT

Airflow

Spark

Description du poste

CACI International Inc. is seeking an experienced AI/ML Platform Engineer in Lisle, IL to support on-prem LLM-enabled technologies and agentic workflows.

You will architect the greenfield implementation, deploy and monitor LLMs on local Kubernetes clusters, and build RAG pipelines integrated with MCP servers. The role requires deep AI infrastructure knowledge, strong DevOps practices, and collaboration across cross-functional teams to iterate improvements and ensure secure, scalable model