Agentic AI Engineer: Vertex AI & Multi-Agent Systems

Capgemini; Nashville (TN); Hybrid; USD 69,000 - 162,000; Full time

Benefits offered by this job

Medical coverage

401(k)

PTO 12-25 days

Life insurance

Employee assistance

Job summary

Capgemini is seeking a hybrid-role AI/ML engineer to design and build agentic AI systems on Google Cloud. You will implement solutions using Vertex AI tools, connect to BigQuery and Cloud Spanner, and develop real-time data pipelines with Dataflow.

The role emphasizes end-to-end problem solving, multi-agent architectures, and production-grade data workflows. A strong background in Python, SQL, and GCP is required.

Qualifications

• Strong experience with Vertex AI and related tools.
• Hands-on GCP work with BigQuery, Cloud Storage, and Spanner.
• Experience in data preprocessing and model deployment.
• Knowledge of multi-agent architectures and real-time processing.

Responsibilities

• Design and build agentic AI systems to automate business processes.
• Develop multi-agent systems for end-to-end problem solving.
• Integrate AI agents with enterprise data sources (BigQuery, Spanner).
• Design and optimize AI/ML solutions on Vertex AI, including training and deployment.
• Build real-time and batch data pipelines using Dataflow and Vertex AI Endpoints.

Skills

Multi-agent systems

Agentic architectures

Real-time processing

Python

SQL (BigQuery)

Tools

Vertex AI

Vertex AI Agent Builder

Model Garden

Vertex AI Pipelines

BigQuery

Cloud Spanner

MCP Toolbox

Vertex Endpoints

BigQuery Vector Search

AlloyDB

Dataflow

Python

SQL (BigQuery)

Job description

Capgemini is seeking a hybrid-role AI/ML engineer to design and build agentic AI systems on Google Cloud. You will implement solutions using Vertex AI tools, connect to BigQuery and Cloud Spanner, and develop real-time data pipelines with Dataflow.

The role emphasizes end-to-end problem solving, multi-agent architectures, and production-grade data workflows. A strong background in Python, SQL, and GCP is required.