Aiyaz Sarwar

Building a Data Science Platform for a pharmaceutical client (BAYER), creating a robust infrastructure tailored to the unique needs of Data Engineers and Machine Learning Engineers.This platform enhances data accessibility, optimizes Machine Learning Model deployment, ensures scalability and security, catering to the unique needs of the pharmaceutical industry.

Roles & Responsibilities:

• Experience in creating, configuring and maintaining Infrastructure on AWS Cloud services including VPC, EC2, SageMaker, S3, Route53, DynamoDB, CloudWatch and IAM using CloudFormation and Terraform as IAC tool
• Executed a project to migrate from on-premises to AWS cloud, resulting in 70% faster deployment time and 45% reduction in infrastructure costs
• Eliminated single point of failure in production architecture designs by deploying service instances (e.g., EC2 or RDS) across two or more isolated AZs based on our customer’s actual resiliency and availability requirements
• Designed and implemented GitHub CI/CD pipeline to automate deployment of AWS infrastructure in Dev, QA and Prod environments
• Implemented security best practices in AWS including multi-factor authentication, access key rotation, role-based permissions, enforced strong password policy, configured security groups and NACLs, S3 bucket policies and ACLs, etc.
• Engineered and maintained highly secure and efficient Kubernetes infrastructure, resulting in 35% increase in system stability and uptime
• Migrated 30+ microservices in Kubernetes (EKS) cluster across 3 regions (EU, US & AP) achieving 60% reduction in time-to-market
• Implemented Airflow, Kubeflow and MLflow to enhance Data and Machine Learning Model processing workflows, achieving a 25% reduction in the Machine Learning development lifecycle.
• Utilized Prometheus, CloudWatch, Athena and Grafana for monitoring and alerting improving system reliability leading to a decrease in incident response time by 65%.
• Lead a team of 10 junior DevOps engineers, providing mentorship and guidance that resulted in a 25% improvement in team productivity.
• Established efficient troubleshooting procedures, reducing ticket resolution time by 50%.

• Designed solution for high availability and business continuity using self-healing-based architectures, failover routing policies, multi-AZ deployment of EC2 instances, ELB health checks, Auto Scaling, and other models based on customer's functional and non-functional requirements.