AI Agent Solutions Engineer

Saatchi US; El Segundo (CA); On-site; USD 97,000 - 149,000; Full time

Job summary

Publicis Groupe is seeking an engineer to shape the engineering standards for an expanding agentic practice. You will own the runtime, flows, connectors and environments for deployed agents and ensure reliability through logging and monitoring.

You will build reusable patterns, implement secure access controls, and drive release discipline with strict versioning and testing. Collaboration with IT and security teams is essential, with strong Python/JSON/YAML skills and a track record of

Qualifications

• 3 to 5 years building and supporting production solutions on the Microsoft stack.
• Built and shipped Power Platform solutions (Power Automate, Copilot Studio) into production inside a governed corporate tenant.
• Dataverse, solutions and environment-based release management.
• Custom connectors and at least one other Azure service in production, such as Azure Functions or Azure Storage.
• Microsoft Entra identity: app registrations, permissions, managed identities.
• Fluent in Python, and comfortable reading and editing JSON and YAML.
• Proven track record of managing competing priorities and delivering outcomes with minimal supervision.
• High intellectual curiosity and ability to rapidly evaluate and apply emerging AI/automation technologies.

Responsibilities

• The runtime: Every shipped agent's flows, connections, solutions and environments, and the answer when one breaks.
• Make failures visible through logging, monitoring and alerts; build safe retry/timeouts/partial-failure handling.
• The plumbing: custom connectors, Azure Functions, document generation, and reads/writes to SharePoint, Dataverse, Excel, Teams.
• Build reusable services, connectors and patterns to avoid duplicating plumbing across agents.
• Release discipline: source control, export/compare, one change at a time, and always a way back.
• Build repeatable tests for flows, integrations and agent behavior, including failure cases and permissions boundaries.
• Design agents with access boundaries and human approvals; maintain a clear activity record.
• Drive work through ambiguity and dependencies, keeping teams accountable to outcomes.
• The doors: data loss prevention, app registrations and security reviews with IT and security partners.

Skills

Microsoft stack

Power Platform

Dataverse

Python

JSON

YAML

Git

Security reviews

Governance

Tools

Azure Functions

Azure Storage

SharePoint

Dataverse environments

Job description

Publicis Groupe is seeking an engineer to shape the engineering standards for an expanding agentic practice. You will own the runtime, flows, connectors and environments for deployed agents and ensure reliability through logging and monitoring.

You will build reusable patterns, implement secure access controls, and drive release discipline with strict versioning and testing. Collaboration with IT and security teams is essential, with strong Python/JSON/YAML skills and a track record of