Title: Agentic AI Engineer AI/ML Engineer
Location: Austin, TX, Warren MI (Onsite)
Duration: 12 months
W2 only
Job Description:
We are looking for a hands-on Agentic AI Engineer Contractor who can design, build, productionize, and support end-to-end AI and agentic systems for manufacturing and enterprise use cases. This is a delivery-focused role for someone who can move from business problem and architecture through implementation, deployment, observability, and ongoing production support.
The ideal candidate brings practical experience across generative AI, machine learning engineering, software engineering, data platforms, AI/ML operations, and enterprise integration. You will work with manufacturing, quality, data, and technology stakeholders to turn operational data and workflows into secure, reliable, explainable AI solutions.
Contract engagement
• Work as an independent contractor on a defined project or statement of work.
• Own technical delivery from discovery and solution design through production release and knowledge transfer.
• Deliver production-ready code, infrastructure, documentation, demonstrations, runbooks, and support processes.
• Collaborate with internal engineering, data, manufacturing, quality, security, and product teams while owning assigned technical deliverables.
What you’ll do
• Design and implement end-to-end agentic AI solutions for manufacturing, quality, operations, and enterprise workflows.
• Build Model Context Protocol (MCP) servers and MCP clients that securely expose tools, data, APIs, and enterprise capabilities to AI agents.
• Develop agent workflows using frameworks such as LangChain, LangGraph, Semantic Kernel, or comparable production-grade technologies.
• Implement agent memory, context management, retrieval, tool calling, structured outputs, planning, and stateful workflows.
• Design and orchestrate single-agent and multi-agent systems with clear roles, handoffs, escalation paths, and failure recovery.
• Integrate agents with enterprise data, APIs, applications, event streams, databases, document stores, and operational systems.
• Use manufacturing and business data to create grounded AI experiences for analysis, investigation, decision support, and workflow execution.
• Apply Databricks, Spark, Delta Lake, lakehouse, vector search, model serving, and related data-platform capabilities where appropriate.
• Build reliable data foundations for AI applications, including data preparation, semantic definitions, metadata, lineage, retrieval strategies, and data-quality checks.
• Implement prompt and response guardrails covering safety, security, privacy, data access, content filtering, tool permissions, hallucination reduction, and human approval.
• Establish AI evaluation practices using representative datasets, golden questions, trace analysis, groundedness, relevance, factuality, task success, latency, cost, and user feedback.
• Productionize AI and agentic systems using CI/CD, versioning, automated testing, environment promotion, infrastructure as code, secrets management, and release controls.
• Implement AI/ML Ops practices for monitoring model and agent quality, data freshness, drift, failures, latency, cost, usage, and operational health.
• Provide hands-on production support, incident triage, root-cause analysis, defect remediation, performance tuning, and continuous improvement for AI and agentic systems.
• Create observability and monitoring dashboards, alerts, traces, audit logs, and operational runbooks.
• Partner with users to validate solutions, incorporate feedback, measure business outcomes, and improve adoption.
• Document architecture, APIs, prompts, tools, agent behavior, evaluation results, known limitations, support procedures, and security controls.
Required qualifications
• Demonstrated hands-on experience designing, building, deploying, and supporting production AI or agentic AI systems.
• Strong Python software engineering skills, including object-oriented design, testing, packaging, APIs, asynchronous programming, and maintainable production code.
• Production experience building both MCP servers and MCP clients, including authentication, authorization, tool definitions, error handling, observability, and secure integration patterns.
• Hands-on experience with LangChain and LangGraph, or equivalent experience with Semantic Kernel or another enterprise agent framework.
• Practical experience implementing agent memory, retrieval-augmented generation, tool calling, structured outputs, context management, and multi-agent orchestration.
• Experience productionizing AI systems through deployment, monitoring, incident response, performance optimization, and ongoing operational support.
• Strong understanding of AI/ML Ops, including CI/CD, model and prompt versioning, automated evaluation, observability, governance, rollback, and environment management.
• Experience designing and implementing prompt and response guardrails for reliable, safe, and policy-compliant AI behavior.
• Experience with AI evaluation and monitoring, including quality metrics, regression testing, trace analysis, groundedness, hallucination detection, latency, cost, and user feedback.
• Working experience with Databricks or a comparable enterprise data and AI platform, including SQL, data engineering, analytics, model serving, or agent-related capabilities.
• Strong SQL skills and the ability to work with structured and semi-structured data such as JSON.
• Experience integrating AI applications with REST APIs, microservices, databases, event-driven systems, identity platforms, and enterprise applications.
• Ability to communicate architecture, technical tradeoffs, risks, and operating requirements to technical and non-technical stakeholders.
Preferred qualifications
• Experience with Databricks Genie Spaces, Mosaic AI, MLflow, Unity Catalog, vector search, model serving, Delta Lake, or similar capabilities.
• Experience with Semantic Kernel, LangSmith, OpenTelemetry, evaluation frameworks, agent observability platforms, or comparable tooling.
• Experience with Azure or another major cloud platform, including containers, Kubernetes, serverless services, managed identities, service principals, networking, and secrets management.
• Experience with manufacturing, vehicle quality, inspection, defect management, production systems, maintenance, supply chain, or root-cause analysis.
• Familiarity with event streaming, message brokers, workflow orchestration, API gateways, and enterprise integration patterns.
• Experience implementing role-based access control, data masking, auditability, privacy protections, and responsible AI practices.
• Experience creating reusable reference architectures, engineering standards, templates, and accelerators for enterprise AI adoption.
Success in this role looks like
By the end of the engagement, you will have:
• Delivered one or more production-ready AI agents that solve validated manufacturing or enterprise problems.
• Built and integrated MCP servers and clients that provide secure, observable access to business tools and data.
• Implemented agent memory, multi-agent orchestration, tool use, and guardrails appropriate to the target workflows.
• Established an evaluation and monitoring framework that makes agent quality, safety, performance, usage, and cost visible.
• Implemented a repeatable AI/ML Ops lifecycle covering development, testing, deployment, monitoring, incident response, and continuous improvement.
• Demonstrated successful production operation and support of AI and agentic systems, including runbooks and knowledge transfer.
• Created reusable architecture and delivery patterns that other teams can apply to additional use cases.
Example use cases
• Answer natural-language questions using governed manufacturing, quality, production, or operational data.
• Summarize trends, investigate issues, and identify relationships across multiple data sources.
• Coordinate multiple specialized agents for analysis, retrieval, validation, recommendation, and workflow execution.
• Connect an agent to enterprise tools through MCP to retrieve data, initiate approved actions, or hand off work to existing systems.
• Generate evidence-backed briefings with links or references to the underlying data and clear confidence limitations.
• Guide users through structured investigations while escalating uncertain, unsafe, or high-impact decisions to a human.
• Monitor agent behavior and automatically identify quality regressions, data issues, integration failures, or unexpected costs.
Working style
• Delivery-oriented: build working systems, not only prototypes or conceptual designs.
• Production-minded: design for reliability, security, observability, supportability, and controlled change.
• Evidence-driven: ground responses in trusted data and make uncertainty explicit.
• Pragmatic: choose the right framework and platform for the use case rather than forcing a single technology.
• Collaborative: work across engineering, data, manufacturing, quality, security, and business teams.
• Continuous-improvement focused: use evaluation results, production telemetry, and user feedback to improve the system.
Suggested job-posting summary
Build and operate the next generation of enterprise agentic AI solutions. As a 360° Agentic AI Engineer Contractor, you will design, develop, productionize, and support AI systems that connect manufacturing data and business workflows to secure, reliable, explainable agents. The role requires hands-on experience with MCP servers and clients, LangChain, LangGraph, Semantic Kernel, agent memory, multi-agent orchestration, guardrails, AI evaluation, monitoring, AI/ML Ops, and enterprise data platforms including Databricks.
