19.10.2026 aktualisiert
Premiumkunde
100 % verfügbarSite Reliability Engineer - Cloud & AI Observability
Wallisellen, Schweiz
Nur Remote
Master's Degree in Management of Companies, Projects Outsourcing and Offshoring - Sofia UniversityÜber mich
I'm a Zurich-based Site Reliability Engineer specializing in Azure and LLM observability. I'm comfortable working with LLMs of all kinds, from instrumentation with OpenTelemetry to Grafana dashboards, chaos testing via APIM and SLO design. AZ-104 certified, with regulated-industry experience.
Skills
SREAzure KubernetesAzure FunctionsAzure DatabricksAzure DevOpsCloud ComputingCloud DatabaseCloud-EngineeringCloud FoundryTelemetrieAzure Active DirectoryKustoAzure AutomationCloud PlatformAzure Data Factory
AI/LLM Observability
Expertise in designing and implementing end-to-end observability platforms for production AI and LLM applications using the LGTM stack (Grafana, Loki, Mimir, Tempo) and OpenTelemetry GenAI semantic conventions to capture prompt/response traces, token usage, and model-level latency.
Cloud Infrastructure & Azure Services
Deep knowledge of Microsoft Azure and Hybrid Cloud environments, including AKS, Function Apps, APIM, Databricks (Unity Catalog), Event Hubs, Storage Accounts, Log Analytics, and Azure Monitor for scalable and resilient cloud operations.
Scripting & Automation
Proficiency in PowerShell, Python, and Kusto Query Language (KQL) for automation, log analysis, diagnostics, and fault injection to support cloud reliability and incident response workflows.
Cloud Operations
Hands-on experience in cloud monitoring, optimisation, and resource management to improve performance, scalability, and cost efficiency across critical banking platforms.
Identity & Access Management
Solid background in Azure AD, IAM, B2B/B2C configurations to ensure secure and compliant access control in regulated enterprise environments.
CI/CD & Infrastructure as Code
Practical experience with GitLab CI/CD pipelines and ARM templates to streamline deployment processes and maintain infrastructure consistency.
Architectural Reviews & Governance
Conducting architectural reviews to ensure compliance with security, resilience, and regulatory standards, and establishing best practices in cloud governance, high availability, and disaster recovery.
Chaos & Resilience Testing
Running chaos and resilience testing against models served through Azure APIM gateways using Python-based fault injection and KQL-driven diagnostics to validate failure handling.
Expertise in designing and implementing end-to-end observability platforms for production AI and LLM applications using the LGTM stack (Grafana, Loki, Mimir, Tempo) and OpenTelemetry GenAI semantic conventions to capture prompt/response traces, token usage, and model-level latency.
Cloud Infrastructure & Azure Services
Deep knowledge of Microsoft Azure and Hybrid Cloud environments, including AKS, Function Apps, APIM, Databricks (Unity Catalog), Event Hubs, Storage Accounts, Log Analytics, and Azure Monitor for scalable and resilient cloud operations.
Scripting & Automation
Proficiency in PowerShell, Python, and Kusto Query Language (KQL) for automation, log analysis, diagnostics, and fault injection to support cloud reliability and incident response workflows.
Cloud Operations
Hands-on experience in cloud monitoring, optimisation, and resource management to improve performance, scalability, and cost efficiency across critical banking platforms.
Identity & Access Management
Solid background in Azure AD, IAM, B2B/B2C configurations to ensure secure and compliant access control in regulated enterprise environments.
CI/CD & Infrastructure as Code
Practical experience with GitLab CI/CD pipelines and ARM templates to streamline deployment processes and maintain infrastructure consistency.
Architectural Reviews & Governance
Conducting architectural reviews to ensure compliance with security, resilience, and regulatory standards, and establishing best practices in cloud governance, high availability, and disaster recovery.
Chaos & Resilience Testing
Running chaos and resilience testing against models served through Azure APIM gateways using Python-based fault injection and KQL-driven diagnostics to validate failure handling.
Erfahrung
Berufserfahrung0 - 5 Jahre
Freelancer seit2026
Sprachen
BulgarischMutterspracheEnglischMutterspracheSerbischMuttersprache
Projekthistorie
Serve as SRE on the Enterprise AI Suite team, building and operating the observability platform for the bank's production AI applications. • Designed and implemented an end-to-end observability layer on the LGTM stack (Grafana, Loki, Mimir, Tempo) covering production LLMs including Granite, Qwen, Gemma, OpenAI GPT, GPT-mini, and Whisper. • Instrumented AI services with OpenTelemetry GenAI semantic conventions to capture prompt/response traces, token usage, and model-level latency alongside infrastructure telemetry. • Built Grafana dashboards and alerting for model health, APIM gateway performance, and per-model cost/latency across the AI Suite's Azure footprint (AKS, Function Apps, APIM, Databricks). • Ran chaos and resilience testing against models served through the Azure APIM gateway, using Python-based fault injection and KQL-driven diagnostics to validate failure handling. • Partnered with AI/ML engineering teams to define SLOs for LLM-backed services and close observability gaps between infrastructure monitoring and AI output quality
Optimised cloud infrastructure to improve performance, scalability, and cost efficiency across critical banking platforms. Conducted architectural reviews to ensure compliance with security, resilience, and regulatory standards. Designed advanced monitoring solutions to enhance visibility, alerting, and incident response. Partnered with development and infrastructure teams to strengthen site reliability and deployment processes. Worked closely with Centres of Excellence (COEs) to establish best practices in cloud governance, high availability, and disaster recovery.
From May 2022 to June 2024 I was working as an External Employee in UBS via Luxoft, After that I was internalised until February 2026.
Monitored cloud environments to ensure stability and performance. Delivered timely incident reporting and proactive issue resolution. Authored and optimised Kusto Query Language (KQL) queries for reporting and troubleshooting. Performed log analysis to detect patterns and enhance cloud service reliability.