Samir Cury

Promoted from within to run the DevOps Organization, spending 30% of time on Internal Strategy, planning and stakeholder conversations, 40% mentoring Leadership and growing the team, and 30% in Customer Facing External Strategy.

Org Size: 16

• Responsible for 3 DevOps/SRE Teams across 5 Timezones, all Operational Concerns, Cloud Infrastructure and implementation of Security, RBAC and VPN for 70+ people in Developer Teams.
• Customer-facing exposure on QBRs, routine strategy meetings, presentations and high-level documents with Senior Executive Leadership.
• Led and shaped the overall DevOps strategy and vision for the organization, aligning it with business objectives and driving technology transformation.
• Implemented deep cultural change in the engineering teams, fostering innovation and modernization in lieu of "legacy works" stagnation.
• Improved DevOps reputation from "blocker" to "enabler" after several programs to enable Service Teams to have autonomy on releasing and in Production, the applicable part of the Google SRE mindset.
• Collaborated with cross-functional teams including development, operations, QA and IoT Device Operations to improve collaboration, streamline workflows, and eliminate bottlenecks.
• Mentored or hired Technical Leadership and Management for the 4 Leadership Functions. 2 were promotions from within as a result of demonstrated growth.

Led the DevOps Enablement Team from Turnaround state to Sustained Success.

Team Size: 4 Reports

40% Implementation+on-call, 60% coordination time.

• Drove the unification of DevOps Platforms into Kubernetes, after thorough architecture improvements and benchmarks, then securing buy-in from Stakeholders/Dev Teams.
• Pioneered Terraform and Kubernetes at Plume, currently running 31 Microservices in 8 Production Environments
• Reduced burnout drastically in the team by simplifying deployment processes and implementing timezone-affinity for assignees.
• Designed and mentored the team to implement a self-serve Framework for New Application Onboarding, removing DevOps as a dependency and scaling Application Onboarding for new products
• Led and managed a team of DevOps engineers to successfully implement and maintain continuous integration and deployment processes, resulting in a 50% reduction in Release deployment time.
• Designed and led implementation of a Terraform Framework that reduced Full Environment Provisioning from 3 to 1 month
• Architected and led modernization of Jenkins from a monolithic, life-support, nearly dysfunctional state, to a distributed, Kubernetes Native model with constant upgrades and improvements.
• Enforced DevOps best-practices in new applications and products as a pre-requisite for Production Launch

Founding member of the AdTech DevOps team, responsible for Infrastructure needs of several teams : Backend, 2 x Data Engineering, Frontend, Data Science.

On a Technical Leadership role, with potential to promotion to management.

• Mapped scope and projects across 5 stakeholder teams and setup DevOps Roadmap
• Hired 3 engineers in 6 months to compose the team
• Introduced state-of-the-art Observability with Prometheus and Grafana
• Led Incident Management best practices in collaboration with Project Managers and Customer Success
• Designed and supported migration of 450+ Data Pipelines from Jenkins to Airflow as a scheduler
• Designed and implemented the Department's migration from on-premise Datacenter to GCP
• Wrote Airflow DAGs for PBs of Data Replication and Consistency checks across DCs
• Main Infrastructure actor on GCP migration of a business-critical set of Data Pipelines which increased productivity by 10x, enabling more business and revenue to the Org.
• Introduced Terraform for Cloud Resource Management and Security Policy enforcement, allowing Developer Teams to self-serve their Change requests through Code Pull Requests.

Focus on SRE, Infrastructure as Code and large scale distributed systems, light Data Engineering involvement.

Strong individual contributor on a distributed team amongst AMER and EMEA. Led a few engineers in discrete projects and efforts.

• Cloud Providers - AWS, GCP
• Developed Terraform Automation for provisioning of Kubeadm Kubernetes clusters
• Benchmarked and launched Ingress Controller for 24k Req/s critical path application
• Developed CI/CD Framework in Gitlab for new Containerized application onboarding
• Trained Developers on CI/CD Framework, supported adoption
• On-call for Production Systems
• Involved in routine deployments or Production migrations
• Common Data Pipeline - Deployed and maintained a 5-region system for ingestion of 92k Events/s via HTTP with Streaming through Kafka and output to ETL Data Lake. Focus on Architecture, Implementation through Cloudera, GCP/GKE + Terraform.
• Co-Leadership on SRE Committee that pioneered a successful Incident Management process, triaging over 50 incidents.
• Main actor and local DevOps lead for AWS->GCP Migration

Solo sysadmin responsible for infrastructure, operations and uptime of www.instructables.com (1,500 req/s in 2015), in addition to participation on more general Microservices DevOps Team

• Introduced Chef as Config. Management for the environment. Replica of production in Dev Workstations.
• Upgraded/Maintained Jenkins for CI/CD w/ Integration tests for 3 codebases and Dev Teams
• Migrated Monitoring system from Cacti to Zabbix, enabling templated Dashboards
• Microservices - Backend in Java Hibernate ; Frontend in Python Django, Javascript Assets
• Infrastructure - Solr as Search Index, self-managed MySQL Cluster, CouchDB Server, HAProxies+Pacemaker, Varnish+VCL for CDN
• Fixed 30% 10s timeouts in Java REST API with Garbage Collection Fine Tuning and Migration from Serial to CMS Collector.
• Designed browser-based telemetry feature for time-on-page tracking, developed backend on NoSQL+Map Reduce and partnered with Frontend Developer for launch.
• Basic AWS Resource Management for minor web applications

Systems administrator of a Bare Metal HPC Facility for particle physics.

Responsible for Compute and Storage resource management, transfer systems, research initiatives in conferences in addition to routine duties.

• Computing - HTCondor Batch System - 5.5k Cores, 14 Racks, 350 Servers
• Storage Systems - Hadoop HDFS - 4.4 PB Raw disk
• Foreman/Puppet for IaaC, OS Provisioning
• HTCondor as a Batch System
• Developed automated CPU Benchmarking system to make data-driven decisions on Hardware Purchases
• Facility went from 7th to 3rd place in the US Ranking of data processing during my tenure.
• 2x more core count purchased per year with similar budget of similar facilities
• Led Network Transfers Benchmarking group amongst 7 sites, after breaking the record of Transatlantic distributed transfer speeds.
• Designed+Implemented updates on Datacenter Power and Cooling systems.

• Phase I - Responsible for Data Reprocessing workflows in 3 regions, 12k cores.
• Operated and Developed multi data-center Workflow Management System (DMWM)
• Phase II - Operated and Developed Workflow Management System for high criticality, 12h SLA Processing system for Accelerator Data.
• Phase III - Technical Leadership for a team of 3 people, interfacing with stakeholders and managers of 12 Physics Groups
• On-call 24/7 AAA for Tier-0 system
• Wrote on-call instructions and commissioned with a pool of 30 on-calls.
• Improved Observability of the system with OSS and integrated with CERN home-grown tools
• Development Tech Stack : Python(+OO/TDD), Oracle DB, Javascript/JQuery, CouchDB
• Published Research in the Journal of Physics regarding compute performance + developed a new workflow modulation algorithm.

• CERN/LHC (Large Hadron Collider) Collaboration
• Daily Operations and expansion projects of the HPC Facility
• Represented the facility in collaboration status updates
• Responsible for dCache & HDFS Distributed Filesystems, HTCondor Batch System and Grid interfaces for Compute and Transfers
• Yearly maintenance of host certificates for 30+ servers
• Datacenter Hardware maintenance
• Networking equipment setup (L2/L3)
• Deployed Xen VM Hypervisor for upgrade tests and prototyping
• Training of student cohorts/new hires
• LDAP System for Local Users and their workstations
• Centralized /home/user directories in NFS
• Rocks Linux / BSD shop