Agentic AI Platform Engineer (with MLOps Expertise)

Prisma Sync Tech; United States; Hybrid; USD 140,000 - 170,000; Full time

Job summary

A technology firm is seeking an Agentic AI Platform Engineer to build scalable, production-grade infrastructure components. The role involves creating standardized AI constructs, focusing on agent orchestration and MLOps practices. Key responsibilities include designing multi-agent architectures, implementing CI/CD for AI agents, and establishing evaluation pipelines. Candidates should be familiar with enterprise-grade MLOps discipline and have experience in AI engineering.

Responsibilities

• Build reusable Agentic components such as MCP servers and data connectors.
• Design multi-agent architectures and tool invocation frameworks.
• Implement CI/CD for AI agents and manage model lifecycle.
• Enable CI/CD for AI agents and manage model lifecycle with MLflow.
• Establish evaluation pipelines, observability, latency monitoring, and security testing.

Tools

SQL

Databricks

NetDocs

APIs

Job description

Role Summary

The Agentic AI Platform Engineer builds scalable, reusable, production‑grade agentic infrastructure components. This role focuses on creating standardized Agentic AI constructs — including MCP servers, connectors, agent orchestration templates, memory frameworks, evaluation pipelines, and deployment patterns — with enterprise‑grade MLOps discipline.

Key Responsibilities

• Agentic Platform Engineering (Build reusable Agentic components):
  • MCP (Model Context Protocol) servers, MCP registries, MCP gateways
  • Data Connectors (SQL, Databricks, NetDocs, APIs)
  • Tool orchestration frameworks
  • Memory & context management services (short term, long term memory)
  • Create standardized agent templates for common agent patterns
• LLM & Agent Orchestration (Design & deploy agents):
  • Multi‑agent architectures and design patterns
  • Design tool invocation frameworks, memory management framework
  • Define & implement guardrails and policy enforcement at runtime
• MLOps & AI Engineering:
  • Implement CI/CD for AI agents
  • Manage model lifecycle using MLflow, model registry, and version control for prompts and agents
  • Establish evaluation pipelines, observability, latency monitoring, hallucination detection, and security testing