Designed and built a new multi account AWS infrastructure using the latest version of Terraform. The code is fully modularised and builds out multiple complete VPCs. IAM Users & Roles are centrally managed through a separate security account and all credentials are secured by encryption at the point of creation via Keybase.

Defined coding standards and coding style for Alchemist Labs own .NET Core 3.1 microservices. Built various Nuget packages to allow all projects to be built in a consistent style. All services are containerised and running in Elastic Container Service (ECS). All projects are automatically built and tested, with the code deployed and Terraform applied by CircleCI.

Completed multiple website projects. All built using the latest versions of React with the latest features such as Hooks and Context API. Websites have used various frameworks including NextJS and Gatsby.

Consulted on Digital Transformation for a large Manchester based e-commerce enterprise. Successfully migrated their mission critical Hybris system and supporting Oracle RAC databases, from IBM SoftLayer to AWS. Introduced new CI/CD pipelines (Jenkins) to ensure the Java micro services could be tested and deployed automatically. All AWS infrastructure was created and managed using Terraform.

Built an entirely new AWS Infrastructure using Terraform. This included 4x different AWS accounts. Infrastructure, Dev, QA & Prod. Was responsible for all aspects of the infrastructure and securing all resources, including configurations for both site to site and point to site VPN.

The first major project was to completely overhaul the internal sales import process. As principal engineer I analysed the existing system, and built a new system on top of the AWS Infrastructure we put in place. The new system (WebAPI hosted in ECS) validated, staged and processed data using AWS Lambda functions in minutes and published sales data to SQS which then pushed data into Elasticsearch. The original system would take hours to process the same data, and was massively prone to errors and impacted performance of all other Ditto systems.

React based web application which displayed customer dashboards with interactive graphs. Data for the graphs was obtained via various API's which talked to a number of different data sources including GraphQL, Elasticsearch and MySQL. The system was massively performant and easily handled accounts of all sizes with data sets of all sizes, replacing an old system which failed to even load for the larger accounts.

Built a logging api and supporting package used in all our applications/services to push all log data to Elasticsearch. Then used Cloudwatch Alarms to monitor various metrics which would raise alarms, publish to SNS and then trigger a series of Lambda functions to push alerts out to various people/places based on severity, including Slack channels/SMS/email.

My 10% time project which allows users to store/retrieve passwords/documents. The system stored documents in DynamoDb so was queryable in the ReactJS based web front end, and all user documents were stored in S3. Both DynamoDb and S3 documents/objects were encrypted by KMS.

Migrated ~30 mission critical legacy services running on a single server. To be independent services, hosted on various parts of Amazon Web Services. This included risky refactoring/rewriting parts of the services, replacing unsupported technologies such as MSMQ with SNS/SQS. The services now utilise many different parts of the AWS offerings, such as Lambda, Elastic Beanstalk, Code Deploy, DynamoDB, Route53, CloudWatch, S3, EC2 to name a few. All services/infrastructure is expressed as Terraform meaning, repeatable deployments with opportunity for human error greatly reduced.

Led the design of a communication service, that enabled the business to have 2 way SMS communication between customers and sales partners for the first time. I used the opportunity to explore AWS Lambda, which was a huge success for the business. Lambda greatly reduced the complexity of the solution. It was faster to build, easier to maintain and financially very efficient. Senior I.T. management have said it’s the most scalable, resilient service we have, and is almost completely free to operate. This was the first Lambda function at Zuto and I presented a brown bag to the rest of the company to explain the implementation and reasoning for using the technology.

I built an entirely new document upload component for our website, that enabled customers to safely upload sensitive documents such as passports, photo id's, finance agreements and other documents in line with new GDPR requirements. I opted to build this using React/Redux which others on our team were not experienced with, but was used elsewhere on the website by other teams. First I built a proof of concept and presented to the team. I then paired with other team members, teaching them React/Redux. Before long, the team was working independently and extending the component until it supported all the various document types.

Re-architected and correctly implemented the in-house CQRS system which I inherited.

Analysed confusing legacy code which had become a myriad of commands calling commands and replaced with a fully distributed Saga utilising the Routing Slip Pattern. Supporting both RabbitMQ and Azure Service Bus message brokers based on environment using dependency injection.

Implemented a centralised OpenID Connect/OAuth2 authentication/authorisation system using IdentityServer. This replaced the existing bespoke login system used in all web applications. Enabling single sign-on for users using Azure AD/Google/Facebook/Twitter etc.

Lead development from inception to delivery of a system, capable of running tasks/applications across multiple servers. Workers were configurable to use parallelisation where necessary. The system provided up to date performance data so that the business could perform valuable data analysis.

Redesigned a legacy data collection system written in Python with an up to date C# equivalent. This drastically improved the speed and reliability when gathering conversion and performance data from 3rd party APIs, and inserting it into client databases. The performance and conversion data was then matched across sources to allow for easier querying and path to conversion analysis. The data was fed into a dashboard to ensure accuracy before presentation to clients.

Designed and wrote a tool to automate the process of trafficking display campaigns into Atlas and DCM ad serving platforms. The tool greatly reduced the time needed to launch business critical advertising campaigns.