WestJet
See all jobs at WestJetSenior Analyst, AI Performance & Reliability
Position Overview:
The Senior Analyst, AI Performance & Reliability is accountable for the health, performance and reliability of AI and machine learning solutions in production. The role monitors AI services and model behaviour, identifies emerging issues, coordinates incident response, and drives improvements that keep solutions stable, available, cost-effective and performing as expected.
The role provides ongoing visibility into service availability, latency, errors, usage, cost, data quality and output quality, including model drift and unexpected or inappropriate results. Working closely with AI / ML Developers, Data Scientists, platform teams, business owners, security and governance teams, the role helps ensure production AI and ML solutions can be trusted to operate consistently and recover effectively when issues occur.
Key Deliverables:
Accountable for keeping AI and ML solutions reliable, available and performing as expected through effective monitoring, incident management, troubleshooting, service recovery and continuous improvement.
Monitor AI and ML solutions in production, including availability, latency, errors, throughput, usage, cost, data quality and output quality.
Build and maintain dashboards, alerts, health checks and reusable monitoring patterns that provide timely visibility into solution health.
Detect model drift, data changes, degraded performance, unexpected outputs, harmful or inappropriate responses and other signs of service deterioration.
Work with business and technical owners to define reliability targets, service expectations, operational thresholds and escalation triggers.
Produce regular operational reporting and communicate material trends, risks and service issues to appropriate stakeholders.
Lead the investigation, triage and coordination of production incidents affecting AI and ML solutions.
Assess business and operational impact, establish incident priority, and coordinate response activities across technical and business teams.
Troubleshoot issues across applications, APIs, models, data pipelines, platforms and infrastructure to identify likely causes and required actions.
Coordinate with AI / ML Solutions Developers and platform teams when rollbacks, configuration changes, model replacements, fixes or redeployments are required.
Validate service recovery and confirm that the solution, monitoring controls and outputs are operating as expected.
Conduct post-incident reviews, document root causes and lessons learned, and track corrective and preventive actions to completion.
Maintain support procedures, runbooks, escalation paths, contact lists and recovery plans.
Identify and address risks that could affect solution reliability, availability, performance, resilience or recovery.
Improve monitoring coverage, alert quality, recovery practices, automation and operational support processes.
Assess production readiness and confirm that appropriate monitoring, support documentation, escalation processes and recovery procedures are in place before release.
Support capacity, performance and cost optimization by identifying constraints, inefficient usage patterns and emerging demand.
Partner with AI / ML Solutions Developers, Data Scientists and platform teams to prevent repeat incidents and improve solution supportability.
Contribute operational requirements and lessons learned to solution design, testing and lifecycle decisions.
Monitor model and AI output performance against defined technical, quality and business acceptance measures.
Identify changes in source data, knowledge sources, user behaviour or business conditions that may affect solution performance.
Analyze feedback, exceptions and performance trends to determine when investigation, correction, tuning or retraining may be required.
Escalate identified model or output concerns to the accountable solution, model, business and governance owners.
Maintain traceable records of identified issues, thresholds, findings and follow-up actions.
Monitor production safeguards and escalate material reliability, privacy, security, responsible AI, model risk or compliance concerns.
Support operational evidence, reporting and documentation required for governance reviews, audits and oversight.
Confirm that access, logging, monitoring and incident-management practices follow approved enterprise requirements.
Help define and improve reliability standards, service-health measures, operational controls and reusable monitoring practices.
Promote consistent incident, problem-management and continuous-improvement practices across AI teams.
Use monitoring, incident, usage, cost and feedback data to identify opportunities to improve reliability, performance and supportability.
Identify recurring issues and coordinate longer-term corrective actions with solution and platform owners.
Share operational insights, lessons learned and emerging risks with technical teams, business owners and leadership.
Contribute to operational reviews, reliability planning, knowledge sharing and the ongoing maturity of AI / ML operations.
Communicate technical issues and service impacts clearly to both technical and non-technical stakeholders.
Experience and Qualifications:
Bachelor's degree in Computer Science, Engineering, Data Science, Information Systems or a related field, or equivalent practical experience.
Relevant certifications in cloud platforms, observability, site reliability, IT service management, AI or machine learning are considered an asset.
4+ years experience supporting production systems, AI solutions, ML platforms, data products or cloud-based services.
Experience with monitoring, observability, logging, alerting and incident-management tools and practices.
Experience troubleshooting applications, APIs, data pipelines, infrastructure or distributed cloud services.
Experience investigating production incidents, coordinating recovery and contributing to root-cause analysis and preventive actions.
Experience using operational data and service metrics to identify trends, risks and improvement opportunities.
Experience working with multidisciplinary technology and business teams in a production support environment.
Understanding of model monitoring, drift detection, data quality, AI output evaluation and changing production conditions.
Understanding of reliability, availability, performance, resilience, capacity, recovery and service-management practices.
Knowledge of service-level indicators, service expectations, operational thresholds and effective alert design.
Ability to analyze logs, metrics, traces, model results and other technical evidence to diagnose complex issues.
Familiarity with cloud platforms, APIs, containers, data pipelines and AI / ML solution architectures.
Familiarity with security, privacy, responsible AI, model risk, governance and audit requirements.
Strong analytical, problem-solving, prioritization and decision-making skills.
Strong communication, coordination and documentation skills, particularly during high-priority incidents.
Ability to work calmly and effectively across a matrixed environment when service issues require rapid coordination.
What Success Looks Like
Production issues are detected early, assessed accurately and resolved quickly.
AI and ML solutions consistently meet defined reliability, availability, performance and output-quality expectations.
Model drift, data-quality issues, unexpected outputs and service degradation are identified before they create significant business impact.
Monitoring provides clear, actionable visibility into solution health, usage, cost and emerging risk.
Incident response, escalation and recovery processes are understood, documented and effective.
Root causes are addressed and repeat incidents decrease over time.
Operational risks and material privacy, security, responsible AI or compliance concerns are escalated promptly.
Business users and solution owners can rely on AI and ML services to operate consistently and effectively.
Operational insights lead to measurable improvements in reliability, resilience, performance and supportability.