AI Engineer

AptivaCorp; Fremont (CA); On-site; USD 96,000 - 152,000; Full time

Job summary

AptivaCorp in Fremont, CA is seeking an experienced AI Engineer to accelerate AI-first engineering across our software ecosystem. You will partner with teams to transform SDLC workflows using agents, LLMs, and Microsoft Copilot while delivering production-ready solutions.

The role emphasizes hands-on engineering, security, observability, and scalable patterns across enterprise platforms. This long-term contract role requires PST availability.

Qualifications

• Bachelor's degree in CS or related field.
• Strong Python programming skills.
• Hands-on experience building AI agents and agentic workflows.
• Experience with LLM tool calls, structured outputs, and orchestration.
• Familiarity with enterprise integrations and REST APIs.

Responsibilities

• Identify high-value engineering workflows for AI-first transformation.
• Design production-ready agentic workflows and reusable tools.
• Integrate agents with enterprise platforms and data sources.
• Ensure security, observability, and cost controls in solutions.
• Document architecture and knowledge transfer for teams.

Skills

Python

LLM/agents

API development

AI integration

GitHub Copilot

Education

Bachelor's degree in CS or related field

Tools

Docker

Kubernetes

REST APIs

Job description

Job Title : AI Engineer

Location : Fremont, CA-Onsite (M-F)

Duration : Long Term Contract-W2

Need Only PST Candidates

Agentic Platform Developer

Role Overview

We are seeking an experienced Software Engineer, Agentic Developer Platform, to accelerate the adoption of AI-first engineering and agentic workflows across our software development ecosystem.

The engineer will partner with engineering teams to identify existing workflows that can be transformed using Generative AI, Microsoft Copilot, agents, and intelligent automation. The focus is on leveraging the Microsoft ecosystem, existing enterprise platforms, APIs, data, and developer tooling to deliver practical, reusable, secure, and production-ready agentic solutions.

This is a hands-on engineering role. The ideal candidate combines strong software engineering fundamentals with demonstrated experience of building, integrating, evaluating, and operating AI agents and orchestrated workflows in enterprise environments.

Engagement Objectives

• Identify and prioritize high-value engineering and SDLC workflows for AI-first transformation using Microsoft Copilot and the Microsoft agentic platform where appropriate.
• Deliver production-ready agentic solutions for selected high-value use cases.
• Create reusable platform capabilities, integration patterns, tools, and skills rather than isolated proofs of concept.
• Put appropriate evaluation, security, reliability, observability, and cost controls in place.
• Document the architecture and transfer of knowledge so internal teams can sustain and extend the solutions.

Key Responsibilities

AI-First Workflow Transformation

• Assess existing engineering and SDLC workflows to identify opportunities for AI, agents, and intelligent automation.
• Redesign workflows using the appropriate combination of LLMs, agents, deterministic software, automation, and human-in-the-loop controls.
• Rapidly prototype, validate, and produce selected use cases based on measurable business and engineering value.
• Prioritize approved Microsoft capabilities and existing enterprise platforms over unnecessary standalone solutions.
• Partner with engineering and business stakeholders to translate workflow challenges into scalable technical solutions.

Agentic Platform Engineering

• Design and build production-ready AI agents and multi-step agentic workflows.
• Develop reusable agent tools, skills, services, prompts, and integration patterns.
• Implement tool and function calling, structured outputs, routing, planning and execution, context management, state, memory, and workflow orchestration.
• Use single-agent, supervisor/worker, planner/executor, routing, and multi-agent patterns where appropriate.
• Implement retries, fallbacks, checkpoints, validation, approvals, and failure-recovery mechanisms.
• Establish reusable architecture patterns that other engineering teams can adopt.

Enterprise & Developer Platform Integration

• Integrate agents with Microsoft 365 Copilot, Microsoft Copilot Studio, Microsoft Foundry, Azure services, GitHub, and other approved enterprise systems, developer tools, APIs, and data sources, as appropriate to each use case.
• Build integrations using Microsoft Graph, Power Platform connectors, REST APIs, SDKs, Model Context Protocol (MCP), events, and other supported interfaces.
• Apply enterprise authentication, authorization, secrets management, data protection, and least-privilege principles.
• Enable agents to safely retrieve information, invoke tools, and perform approved actions across Microsoft and other approved enterprise platforms.

AI-First Software Engineering

• Use GitHub Copilot and approved Microsoft Copilot experiences throughout the development lifecycle, supplemented by comparable approved AI-assisted engineering tools when appropriate.
• Apply AI assistance to design, coding, refactoring, testing, debugging, documentation, and troubleshooting.
• Identify opportunities for agentic automation across requirements, development, code quality, testing, CI/CD, documentation, operations, and engineering analytics.
• Demonstrate practices that can be adopted by the broader development organization.

Microsoft Copilot & Agentic Platform

• Design solutions that use Microsoft 365 Copilot and Microsoft Copilot Studio when low-code orchestration, enterprise knowledge, actions, and user-facing copilot experiences are the best fit.
• Use Microsoft Foundry and Foundry Agent Service for code-first, model-flexible, governed, and observable agent workloads when appropriate.
• Build or consume reusable tools through MCP, Microsoft Graph, Power Platform connectors, Azure Functions, Logic Apps, and secured enterprise APIs.
• Implement connected agent and multi-agent patterns across Copilot Studio, Foundry agents, and specialized enterprise agents when they improve modularity and maintainability.
• Integrate Microsoft Entra ID, managed identity, role-based access, telemetry, evaluations, and guardrails into the production design.
• Publish agent experiences to Microsoft 365 Copilot, Teams, or other approved channels based on user needs and enterprise governance.

Evaluation, Reliability & Observability

• Establish automated evaluation frameworks and representative test datasets for AI and agent workflows.
• Evaluate grounding, retrieval quality, tool-call correctness, structured-output accuracy, workflow completion, and regression behavior.
• Implement tracing, logging, telemetry, monitoring, and troubleshooting across multi-step workflows.
• Measure and optimize quality, reliability, latency, token usage, and cost.
• Build automated unit, integration, workflow, and AI evaluation tests and incorporate them into CI/CD.

Required Qualifications

• Strong professional software engineering experience building and supporting production applications and services.
• Strong programming skills in Python and modern backend or API development.
• Hands-on experience building applications using Generative AI and large language models.
• Demonstrated experience developing AI agents and agentic workflows.
• Experience with LLM tool or function calling, structured outputs, orchestration, context and state management, and multi-step execution.
• Experience with retrieval-augmented generation (RAG), grounding, embeddings, and enterprise information retrieval.
• Familiarity with MCP or comparable agent-to-tool integration approaches.
• Experience with LangGraph or an equivalent agent and workflow orchestration framework.
• Strong experience with REST APIs, JSON, Git, automated testing, and CI/CD.
• Experience deploying and supporting cloud-based applications.
• Understanding application security, authentication, authorization, secrets management, and least-privilege design.
• Experience with production logging, tracing, observability, monitoring, and troubleshooting.
• Hands-on experience using GitHub Copilot or comparable AI coding tools in day-to-day engineering.

Preferred Qualifications

• Hands-on experience with the Microsoft Agentic Ecosystem, including Microsoft 365 Copilot, Microsoft Copilot Studio, Microsoft Foundry, Foundry Agent Service, or Microsoft Agent Framework.
• Experience building and publishing enterprise copilots, custom agents, connected agents, or AI-powered developer tools for Microsoft 365 Copilot, Teams, or other enterprise channels.
• Experience with multi-agent architecture and orchestration.
• Experience with Microsoft Agent Framework, Semantic Kernel, or similar agent development frameworks.
• Experience with vector databases, semantic search, and enterprise RAG.
• Experience integrating GitHub, Jira, Confluence, and enterprise data sources.
• Experience with Docker, Kubernetes, and cloud-native architecture.
• Knowledge of enterprise identity and access management, including Microsoft Entra ID, managed identities, and role-based access control.
• Experience implementing human-in-the-loop approval, escalation, and governance patterns.
• Experience optimizing LLM applications for quality, latency, reliability, and cost.
• Workflow assessment and prioritized backlog of AI-first opportunities.
• Production ready agentic solutions for selected high-value use cases.
• Reusable agent tools, skills, services, and enterprise integration patterns.
• Reference architecture covering security, identity, data access, orchestration, and human approvals.
• Microsoft platform implementation blueprint identifying when to use Microsoft 365 Copilot, Copilot Studio, Microsoft Foundry, Foundry Agent Service, MCP, Microsoft Graph, Power Platform, and custom code.
• Evaluation, automated testing, observability, and cost-monitoring practices.
• Technical documentation, operational runbooks, reusable examples, and engineering standards.
• Knowledge-transfer sessions and hands-on enablement for internal engineers.
• Measurable outcome summary, such as reduced manual effort, improved cycle time, increased automation, or improved developer experience.