Kubernetes + GitOpsAs a SRE, I...- provide guidance on engineering effectiveness- maintain logging and monitoring- build tools like helm charts for common deployment patterns- deploy open source operators for performance testing, CI/CD, monitoring, service mesh, routing- build custom operators in golang/python to automate company specific processes- debug infrastructure issues- work with developers to deploy apps and troubleshootAccomplishments:- Promoted from Senior to Principle SRE- Built and maintained a Kubernetes operator that eliminated the complexity and burden of alerting and paging.- Migrated all CI from Jenkins to Github Actions and wrote reusable pipelines that push the boundary of what is capable with Github Actions.

Having primarily worked with AWS ECS before; here I developed competency in writing Kubernetes manifests (Deployments, Pod spec, etc).I was situated in a position where I could mentor more junior engineers while being mentored by the more senior architects from Candid. Even though I was new to Kuberentes, I still surprised the client and my peers, with automation and rigorous testing.Achievements:- Routinely debugged our Gloo Gateway for misconfigured virtual services and remediated them.- Contributed to terraform config to stand up EKS (AWS Managed Kubernetes), and bootstrap initial services like Flux/Prometheus/Etc- CI/CD for Kubernetes config changes, including baking images for the Kubernetes workers.

Maintenance of a large home-grown CMS. Mentoring of junior DevOps engineers. Designed scopes of work, created actionable stories and shepherded them to completion.

Enterprise Jenkins TeamI worked on the team providing Jenkins for all of Capital One (70+ Jenkins masters, 7000+ users). In this position I played a major role in migrating Jenkins agents from a nearly static EC2 based architecture to ephemeral ECS based agents. Essentially each bulid gets a unique, disposable jenkins agent as a docker container. This overall increased platform stability and saved approximately 7 million dollars a year in AWS infrastucture costs by provisioning small (30-60 node) multitennant ECS clusters, agnostic to the workloads that were being run on the cluster.To accomplish this goal I wrote almost half of the Terraform, python library, and CLI used to control and deploy the new infrastucture. I worked with users to discover use cases, remediate bugs, and wrote new tests to future proof against regressions. Technical challenges included: isolating user space docker builds from the host infrastructure (docker daemon as a sidecar container, instead of host daemon), applying aws privileges (AWS IAM) and networking (VPC networking, security groups) at the container level, and the overall challenge of meeting existing user expectations in the new architecture.

Digital Messaging Platform:Actively tried to change Ops heavy DevOps culture in a group that uses a large ECS and Message Bus architecture in AWS. To that end I maintained and improved software pipelines using infrastructure as code (Terraform) and strong integration testing (e.g. Python/Pytest).Achievements:- Provided support for and maintained ECS Clusters (8+ clusters, 100+ developers, 100+ services).- Routinely deployed infrastructure and applications using blue/green and canary style deployments.- Wrote a python library for the automation of Terraform import of existing resources and the programmatic generation of the configuration of those resources. This functionality currently does not exist in Terraform and allows users to clone the configuration of entire ECS clusters (AWS resources include: Route53, ALB/ELB, Listener Rules, Target Groups, ECS Tasks, ECS Services, EC2 instances, ASG's and any additional resources). Providing the ability to deploy new clusters from existing ones or back up cluster configuration in the event of catastrophic failure.- Mentored rotational trainees from the Capital One Technology Development Program- Served as the subject matter expert on the team for Terraform, Docker, and Python

I help clients modernize and employ best practices: Git, Docker, Jenkins, CI/CD, Amazon Web Services (AWS), Agile, Service Monitoring/Metrics, Load Testing, Data Backups. I also supplement client expertise by supplying infrastructure as code.Code Rescue:Migrated legacy PHP/Angular1 applications from "by hand" deployment to Docker and commit triggered continuous deployment with Jenkins to AWS Elastic Container Service. Dove into client code to refactor and implement new features on tight deadlines at the client's request. Documented previously unknown and scattered development practices and infrastructure in Git and Confluence. Acquired necessary domain knowledge by working closely with stakeholders. Handled operations tasks related to domain name, SSL/TLS, data loads, database maintenance/migration, setting up development/production environment, coordinating production releases.ETL:Automated ETL of legacy databases from database dumps and Box API, converting to CSV and modern databases (PostgreSQL/AWS Redshift) with python.Monitoring:Deployed monitoring software (DataDog) to production docker swarm cluster in AWS, to integrate with and collect metrics on services such as MongoDB, RabbitMQ, Nginx, Elasticsearch, and EC2 host level metrics. Created dashboards in DataDog console, providing easy to consume, time series, graphical output of service critical metrics. These were used to analyze performance, scaling, and prevent system failures.

MAXX Potential provided the crucial opportunity for me to retool and retrain after my time in academia. I quickly progressed from software development initiate to junior developer. The variety of work here exposed my interest in DevOps.Large Wordpress Multi-site (4 months)- Fulfilled feature requests submitted by Content Managers in JIRA- Styling and templating using HTML/CSS/JS/PHP- Plugin management and custom plugin editing (PHP)- Automated testing with CasperJSSelenium Automated Testing (2 weeks)- Piloted automated testing using selenium for a client.- Wrote Selenium tests using NodeJS to test a single page React app, including, changing search filters, and inserting search terms and verifying results.Operationalized a Machine Learning Model with Python/Flask (6 weeks)- Tasked with making machine learning predictions available via a REST API- Created a Flask API endpoint that ingested single records and returned an overall risk factor.- Used Sci Kit Learn's gradient descent model to produce weights for the relationship between risk factors and overall risk assessment.- Used Sci Kit Learn's Pipelines to standardize transformation and normalization between Prediction and Training phases.- Used multiple Sci Kit Learn transforms including one hot encoding.- Used Pandas for data manipulation.- Serialized and persisted the trained model for use during prediction (Pickle/Dill).Supported a Development and Operations contract (6 weeks)- Automated Mongo and Jenkins backups using Jenkins.- Complex BASH scripting.- AWS instance and security group management.- Deployed microservices with Docker and Docker Swarm.- Created custom docker images from Dockerfiles.- Created new base OS images from AWS AMIs for use with Docker.

Studied neuronal mechanisms contributing to the development of epilepsy. Used Linux and Matlab to automate signal analysis and feature extraction of neuronal data. Performed statistical analysis and interpretation of large data sets resulting in multiple co-authored publications.

Studied the role of Lacritin, a protein found in human tears, and its possible use in the treatment of chronic dry eye. To this end, I performed weekly production and purification of a recombinant Lacritin protein from E.Coli (including growth, induction, lysis, column chromatography, quantification assays, storage, etc.).