The AI Platform Operations Support role is responsible for enabling the reliable, observable, secure, and governed operation of AI, machine learning, data products, and enterprise agents across the organisation.

This role supports AI, Machine Learning, Data, and business-led agent delivery teams by implementing and operating MLOps, DataOps, DevOps, and AgentOps practices, along with comprehensive logging, monitoring, telemetry, product analytics, request tracking, and post-deployment support processes. The role ensures that AI solutions and enterprise agents are production-ready, operationally stable, cost-aware, supportable, and aligned with enterprise cloud architecture, security, and governance standards.

The role acts as a key bridge between AI/Data delivery teams, the Center of Practice, business requestors, IT Infrastructure, Cloud Architecture, Security, Governance, and Platform teams, ensuring AI products and enterprise agents scale safely and sustainably beyond pilots.

This is not a product ownership role and not a pure infrastructure role. It is an operational enablement and support role focused on observability, request intake coordination, production support, quality, governance alignment, and operational sustainability.

The purpose of this role is to ensure that AI products, data products, and enterprise agents are supported by strong operational foundations rather than ad-hoc or fragile production setups.

This includes supporting the operational flow of enterprise agent requests from intake through deployment and ongoing support; owning operational logging, usage telemetry, metadata management, monitoring, analytics, and quality controls across AI pipelines, platforms, and deployed agents; and ensuring alignment with enterprise cloud architecture, governance policies, access controls, and cost-management practices.

Key responsibilities

1.

- AI, ML, data, and enterprise agent operations support
- Provide operational support for AI, ML, data products, and enterprise agents across development, non-production, and production environments
- Monitor pipelines, services, workflows, applications, and deployed agents for failures, degradation, abnormal behaviour, or support issues
- Support incident triage, investigation, root-cause analysis, and resolution coordination
- Work with delivery teams, business owners, and the Center of Practice to improve operational stability and reduce recurring issues
- Ensure production support expectations are clear before AI products or agents are released to users

2.

- Enterprise agent request management and operational coordination
- Support the intake, tracking, triage, and operational coordination of enterprise agent requests submitted by business teams
- Maintain visibility into the agent request pipeline, including request status, priority, dependencies, risks, target users, and expected support model
- Coordinate with AI Product Owners, Business Partners, the Center of Practice, Security, Data Governance, and Platform teams to ensure requests follow the approved operating model
- Validate that required information is captured before build or deployment, including business owner, intended users, approved data sources, access model, risk level, expected value, and support requirements
- Help distinguish requests requiring simple Copilot enablement, Copilot Studio configuration, workflow automation, or more advanced Azure/Foundry engineering support
- Ensure agent requests are not treated as isolated builds, but as governed digital products with clear lifecycle ownership and operational accountability

3.

- MLOps, DataOps, DevOps, and AgentOps practices
- Support establishing and executing MLOps, AIOps, DataOps, DevOps, and AgentOps practices aligned with enterprise standards
- Support model, data pipeline, and agent deployment, versioning, rollback, lifecycle management, and controlled promotion across environments
- Enable repeatable, auditable deployment and operational patterns for AI products and agents
- Reduce reliance on manual, fragile, or non-scalable operational processes
- Maintain operational runbooks, support checklists, and release-readiness artifacts for recurring AI and agent deployment patterns

4.

- Logging, monitoring,



observability, and agent telemetry
- Own operational logging across AI, data, and agent platforms, including pipeline execution logs, application and system logs, model inference and performance logs, agent conversation and action telemetry, data quality and validation logs, and support ticket or request history where applicable
- Maintain log tables, metadata, and telemetry to support traceability, diagnostics, auditability, and continuous improvement
- Implement monitoring dashboards and alerting for system health, reliability, performance, adoption, agent usage, failure rates, and escalation patterns
- Ensure logs and telemetry align with enterprise standards, privacy expectations, access controls, and governance requirements
- Provide operational visibility into which agents are being used, by whom, for what purpose, and whether they are delivering expected value within approved boundaries

5.

- Model, data, agent, and quality assurance support
- Monitor AI model performance, stability, drift, and quality indicators in production
- Support validation and quality checks for models, data pipelines, workflows, prompts, knowledge sources, and deployed agents
- Ensure testing, security, access, and quality gates are enforced before production releases
- Partner with AI developers, data scientists, business analysts, and the Center of Practice to improve robustness, reliability, and user experience
- Help identify recurring agent failures, weak knowledge grounding, missing permissions, misunderstood user intents, broken automations, or poor escalation paths

6.

- Post-deployment agent support and continuous improvement
- Provide first-line or coordinated second-line operational support for deployed enterprise agents, depending on the agreed support model
- Triage user-reported issues such as access problems, broken workflows, unexpected answers, missing knowledge sources, performance degradation, or failed handoffs
- Coordinate fixes with the appropriate owner, including the Center of Practice, AI developers, business owners, platform teams, data owners, or vendors
- Maintain support documentation, known-error logs, FAQs, release notes, and operational runbooks for deployed agents
- Analyze post-deployment feedback and usage telemetry to recommend improvements, retirement, consolidation, retraining, or escalation of agents
- Ensure agents remain current when source documents, business processes, policies, integrations, or access models change

7.

- Product analytics and usage telemetry
- Implement product analytics and usage telemetry for AI solutions and enterprise agents, including usage patterns, adoption signals, performance metrics, user satisfaction indicators, operationalindicators, cost indicators, and support demand
- Provide insights to Product Owners, business owners, Center of Practice leads, and stakeholders to support value tracking and optimisation
- Enable transparency into how AI products and agents are used and perform over time
- Identify low-adoption, high-risk, duplicative, or high-support-demand agents that require intervention

8.

- Cloud architecture, governance, and compliance alignment
- Ensure AI, data, and enterprise agent solutions align with enterprise cloud architecture standards
- Work closely with Cloud Architecture, Infrastructure, Security, Governance, Data, and Center of
- Practice teams
- Support compliance with security, access control, data-handling, responsible AI, and approved-source policies
- Provide operational evidence for governance reviews, audits, and agent portfolio reviews
- Help ensure deployed agents remain within approved security, data access, privacy, retention, and auditability boundaries

9.

- Cost, capacity, and resource monitoring
- Monitor cloud resource utilisation across AI, data, and agent workloads




- Support cost optimisation and capacity-planning efforts
- Provide visibility into trade-offs between performance, scalability, reliability, user adoption, and cost
- Flag inefficiencies, abnormal consumption, duplicated agents, or support risks proactively

10.

- Release readiness and production enablement
- Support production readiness and go-live reviews for AI products, data products, and enterprise agents
- Ensure operational requirements are met prior to wider deployment, including monitoring, ownership, support path, access controls, release notes, rollback approach, and user communication
- Enable smooth handover from development into ongoing support
- Improve confidence in scaling AI solutions and enterprise agents beyond pilot stage
- Confirm that each deployed agent has a business owner, technical owner, support model, telemetry baseline, and lifecycle review cadence

Qualifications

Education

- Bachelor’s degree in computer science, engineering, information systems, data analytics, or a related field
- Equivalent practical experience is acceptable

Experience

- 5+ years of experience in platform operations, DevOps, DataOps, MLOps, cloud operations, enterprise application support, or AI operations roles
- Experience supporting production analytics, data, AI, automation, or enterprise agent systems
- Experience working with enterprise IT, cloud, security, governance, or platform teams
- Experience with Microsoft Azure and Azure-based services
- Experience supporting analytics and AI workloads using platforms such as Microsoft Fabric or equivalent
- Experience supporting Microsoft 365, Copilot, Copilot Studio, Power Platform, Azure AI, or equivalent enterprise agent and automation platforms is an asset
- Experience coordinating support queues, request intake, production incidents, release readiness, or service management processes is an asset

Technical knowledge

- Strong understanding of MLOps and model lifecycle management
- Strong understanding of data pipelines and data quality monitoring
- Strong understanding of DevOps, CI/CD, release management, and production support practices
- Familiarity with logging, monitoring, observability tooling, metadata management, telemetry, and product analytics
- Familiarity with enterprise agent patterns, prompt/knowledge-source governance, access controls, workflow automation, and post-deployment agent support
- Familiarity with governance and security patterns in enterprise cloud environments

Skills

Strong diagnostic and problem-solving skills

Ability to operate across technical depth, service management discipline, and enterprise constraints

Clear communication with both technical and non-technical stakeholders

Strong documentation and operational discipline

Ability to coordinate requests across business teams, Center of Practice resources, platform teams, security, governance, and vendors

Ability to translate user issues into clear operational tickets, root-cause themes, improvement actions, and release-readiness requirements

Personal attributes

Detail-oriented and pragmatic

Structured and reliable under operational pressure

Strong ownership and accountability mindset

Collaborative and service-oriented

Comfortable supporting business users while maintaining disciplined governance and operational controls

Continuous-improvement mindset focused on making AI and agent operations repeatable, scalable, and safe.

Success measures

Success in this role may be measured through:

Operational stability and reliability of AI products, data products, and enterprise agents

Quality and consistency of logging, telemetry, monitoring, and support documentation

Clear visibility and disciplined tracking of enterprise agent requests from intake to deployment and support

Reduction in operational incidents, recurring agent issues, and manual interventions

Improved visibility into usage, adoption, performance, cost, and support demand

Improved speed and quality of post-deployment support for enterprise agents

Increased confidence in scaling AI solutions and enterprise agents into production

Evidence that deployed agents remain governed, supportable, current, and aligned with approved data sources and enterprise standards

#J-18808-Ljbffr

📌 AI Platform Operations Support (Medellín)
🏢 Kruger
📍 Medellín