Lee Oades

Developed a new enhanced search capability for the UK Government's Explore Education Statistics system.

Introduced an event-driven mechanism (Event Grid events to Storage Queues) to provide the new functionality without further coupling.

I created multiple simple Azure Function apps to consume the events and generate, update or delete content blobs in Blob Storage for the search indexer to consume.

Introduced Serilog to provide structured logging into Application Insights.

Provided technical leadership, training and mentoring to the Hive IT developers on software practices, design improvements and unit testing.

To view examples of my work, this project is open source: https://github.com/dfe-analytical-services/explore-education-statistics/

Maintained and enhanced features in a complex, asynchronous, low latency, event sourced / event driven betting platform.

Built a football multi-game simulation component to emit game events (goals, cards, extra time, penalties) and betting events in a reasonably realistic manner. This is used for both load testing and simulation of rare events e.g. penalty shootouts.

Developed a new mechanism in the betting platform to handle how liabilty is grouped and aggregated for related markets to better coordinate the pricing to reduce opportunities for arbitrage.

Implemented several autonomous components to encapsulate behaviours and execute actions based on events, timers and the state of the system.

Made heavy use of Actor architecture using Microsoft Orleans that consumer Kafka streams. Communication with external components via gRPC streams and Swagger end points. Persistence to MySql or DynamoDB where appropriate.

High test coverage using XUnit and Specflow tests.

Utilised many technologies and tools: Docker, K3D, Tilt, Kubernetes, Kowl, RedPanda, OpenLens, Helm, Kibana, DataDog, plus many AWS services

My use of the state machine `Stateless package within Orleans revealed a bug in the library which I reported and fixed. The fix was accepted and merged into the library.

Designed and built the middleware layer to interface the data, actions and events between the front end Magento system and numerous custom Builders Merchant ERP (Enterprise Resource Planning) systems.

Beautiful thin vertical slice components to better isolate feature implementations. Use of Mediatr for clean code separation.

Event driven architecture within AWS - API Gateway endpoints emit events into SNS, queued into SQS, and handled by single responsibility Lambdas. Persistence in MySql and/or DynamoDB.

Declared the infrastructure and deployed it using Terraform.

Candidate registration system : Multiple independent, asynchronous .NET Core microservices, APIs, ASP.NET MVC websites and a Blazor client app deployed to AWS.

Distributed, event driven architecture within AWS (SNS, SQS, Lambdas, DynamoDB).

Cleanly built components utilising Mediatr, built using TDD, XUnit and mock builders.

Also co-developed .NET Core web application to replace legacy system. Included simple isolated SPAs in Knockout and Typescript, dynamic creation of Excel and Word reports using business supplied templates, sending customised emails.

Developed .NET Core backend services for a new, large and complex distributed system for Mortgage Consultants to process new mortgage applications.

Hosted in Docker containers using Kubernetes.

Event driven architecture over Kafka. Storage in MongoDB.

Integration with many existing legacy services.

Built .NET Core ASP MVC url for exam invigilators - Knockout, Bootstrap, OAuth, SQL DB, deployed to Amazon AWS EC2

Developed .NET image processing algorithms for detecting answer from a photo of an exam paper, eliminating lighting abnormalities / shadows and determining which circles were coloured in. High profile project request from CEO, concept to delivery in 3 months.

Significantly improved customer experiences and replaced several tedious, internal, manual processes through automation. Used AWS Lambdas and Step Functions to integrate internal systems with 3rd party vendors, Salesforce APIs.

Implemented a complete workflow for Salesforce development team – introduced source control, continuous integration, automated deployments, environment specific configuration. Provided training, documentation and support.

Front Office FX application provides live + historical data for quote and deal blotters, dealer intervention, position display.

WPF application.

Low latency streaming (RV & Solace into RX)

High unit test coverage and reduced regression costs through BDDs (nBehave).

Back office P&L / Balance Sheet sign off application.

Multithreaded WPF application making heavy use of RX for its multithreading, receiving of updates, UI interaction etc.

Well-architected code and separation of concerns through MVVM, view-model first.

TDD maintains very high test coverage (+90%).

Flagship Commodities Trading Platform

Silverlight v5 multi-window out-of-browser application for traders providing live market prices.

Written in C# and F#. Created new features within the application as well as maintaining it, refactoring and improving its architecture, stability and performance, plus troubleshooting and fixing issues.

Autofac, MEF, RX, TDD, Moq, AgUnit, Silverlight Unit Testing Framework, SpecFlow, interop to C++ COM and Win32 components.

Agile methodologies, pair programming, peer reviews. Provided technical support to team members.

Developed enhanced WPF application framework including layouts, state persistence, magnetic windows, cross-process communication and eventing, common controls and styling and more

Live FX Trades blotter WPF windows application that displays live, filterable grids of FX trade data. TOC subscription service, communication using Protobuffers through RV layer, application used RX to throttle incoming data.

Live FX Position monitor WPF app that displays live FX positions on animated bar charts. Data can be drilled into to display different cuts e.g. by currency pair, client sector, region etc. Windows service uses continuous queries onto Coherence database and MDX queries into OLAP Cube.