William Winter

• Served as Solutions Architect and Technical Lead for a critical Generative AI tool, enabling efficient management of incidents during Olympics streaming on Peacock. Delivered project from concept to MVP in 8 weeks, showcasing strong stakeholder communication up to Senior Vice President level.
• Led solutions architecture for a new Event Management Platform, overseeing architecture and engineering decisions for 3rd party contractors. Ensured alignment with company standards and best practices, resulting in seamless integration and improved operational efficiency.
• Implemented Application Performance Monitoring (APM) across 50+ microservices, providing end-to-end tracing and identifying bottlenecks. Re-engineered API Gateway in Go from NodeJS, reducing response time from 2s+ to under 200ms.
• Led standardisation initiatives across 50+ microservices, reducing average ticket resolution time from 3 weeks to 1 week, significantly enhancing development velocity and team productivity.
• Optimised logging practices, saving the team over $240,000 annually, demonstrating strong FinOps understanding and cost-optimisation skills.
• Enhanced developer productivity through code standardisation initiatives and maintenance of Terraform pipelines for frontend and backend applications.
• Provided mentorship to junior team members, accelerating their career growth and fostering a culture of continuous learning.

• Played a pivotal role in developing a new headless content delivery platform built on WordPress, utilising Symfony, Docker, and AWS technologies (Lambda, CodePipeline, ECS), demonstrating proficiency in modern web architectures and cloud services.
• Engineered and maintained a new ElasticSearch 7 service and associated PHP microservice, showcasing expertise in search optimisation and microservices architecture.
• Designed and implemented a related content module leveraging ElasticSearch similarity search, significantly enhancing reader engagement and content discoverability, driving measurable improvements in user interaction metrics.
• Orchestrated the migration to Docker-based development environments, replacing Vagrant, which streamlined local setup processes and ensured local/production parity, resulting in substantial improvements in developer productivity and deployment consistency.

Led the development of an AI chat assistant using Retrieval-Augmented Generation (RAG) architecture. Designed and implemented a scalable solution deployed on Azure Container Apps, integrating multiple data sources and leveraging Azure AI Search. Architected custom Python components for advanced query processing, multi-index search, and dynamic response generation using GPT models. Directed the development team, created the project roadmap, and managed stakeholder relationships at the executive level. The system significantly improved operational efficiency and decision-making processes across the organization.

In preparation for a high-profile NFL game on Peacock, I led a performance optimisation initiative. We implemented APM across 50+ microservices, identifying a NodeJS API gateway as the main bottleneck. By rewriting it in Go, we reduced request latency from over 2 seconds to 200ms. I also introduced Smoke Testing with response time checks for ongoing performance monitoring. The project utilised Go, Prometheus, Grafana, Tempo, and OpenTelemetry.

To improve my Go skills, I created a Discord bot that tracks user activity in voice chats and assigns roles based on participation. This project helped me understand event-based architecture by implementing listeners for Discord events. I also built a dashboard using PHP (Laravel framework) and TailwindCSS, allowing users to configure the bot for their Discord servers. The project utilised Go for the bot's core functionality, with PHP, MySQL, and Laravel for the web interface.