Worked on team implementing recruiting system for US Army. System constructed using Python in conjunction with AWS serverless components including AWS Lambda and AWS SQS/SNS. Combination serverless and kubernetes container framework.

AWS Aurora Postgres- Worked on logging back-end where database logs are stored and sent to the logging storage system. These logs where then parsed and the result of the transaction persisted to aurora storage.- Developed postgres extensions for management of internal postgres information.AWS RDS on VMWare- Tech lead on product that managed customer owned databases on their on vSphere cluster. This enabled customers to get data within their data center.- Worked directly with VMware on snapshot mechanism for database backup and logs.- Worked with VMware for storage of customer data, VSAN.AWS Kinesis:-Member of team in Amazon MSK launch to ensure control plane work is idempotent. The control plane is a server-less architecture. As well as using Lambda (sever-less)-Member of team to launch AWS Managed Streams for Kafka. A managed service offering kafka to customer with little to no operational challenges. MSK is also fault tolerany, self healing on hardware, network, and data center (availability zones).-Designed AWS Glue Schema Registry which launched in late 2020. The design was a server-less approach to speed time to release.-Designing components of a yet to be released component. A high speed, redundant, and durable service involving streaming.

-Implemented parts of automated email system. Included securing the emails and ensuring automated email generation was legal. Used DKIM to ensure email was not tampered with during send. Ensure a high sender reputation so that email does not end up in spam folder by following email best practices.-Implemented part of Salesforce's group messaging board. Analyzed signals to rank replies to question based on previous user responses and the number of likes per message.

Joined AWS In Memory Team, January 2014-Designed auto-failover solution for Elasticache's Redis offering. Took product from inception to production. Drove implementation, testing, and operations leading a five member team. Architecture used a consistent-hash to slice up instances to be monitored for a fleet of health checkers. It also used an AWS Paxos system to ensure one health monitor watched one instance. This was to ensure on a host outage each instance was still monitored by only one health checker.-DynamoDb, a highly scalable NoSQL Amazon Web Service. Service provides low millisecond response time for inserting and querying for data. System scales as needed to very large data size and throughput. Focus on automatic hardware failure scenarios, fast manipulation of increasing table size, throughput, and durability. Developed task-engine for control plane to handle scaling, creating and modifying dynamoDb tables.-Core team member of RDS, Relational Database Service for AWS. Primary design, lead, and implementation of key aspects of distributed monitoring system. -Delivered Multi-AZ feature of RDS which replicates data between two distinct geographic regions for increased data durability. Also provides for fast failover when a database server fails. Designed and implemented services to monitor tens of thousands database instances and respond to instances needing failover, backup, and software upgrades.

Large scale ad delivery system for various clients. System provided targeted advertisements based on user's demographics and web site browsing characteristics.

Developed a high-speed in house web-server for processing requests for Amazon Web Services. System interfaced with other components to compute billing, check authentication, detect denial of service attacks, and allow a configurable pipeline to individual services to handle their specific needs.Developed release plan for Amazon's Simple Queue Service, SQS. SQS is a message system that allows various types of readers and writers to communicates across disparate servers, networks, languages, and OS types. Successfully released SQS as product still successfully in production today.

Key member of mail delivery team. Implemented system which adds in determining an email spam index. In particular investigate the user's social network and determine if the user is linked to the sender by some degree of separation.Designed and implemented system address book system for entire aol membership. Also implemented an extremely high throughput system to automatically added mail sent from a member to a another member as a contact.Developed system to allow multiple services to communicate with various databases using a set of related servers. System would reduce load on particular databases while allowing requests to remain at a large rate.