About - Matthew van Bird
Staff engineer, 14 years building .NET systems on AWS - APIs, queues, distributed workloads, behavioural data platforms and real-time personalisation. Based in the UK.
Selected work
Behavioural data lakehouse: designed and led a first-party behavioural analytics platform - Snowplow event collection into a Snowflake lakehouse, modelled with dbt in a medallion architecture, orchestrated with Orchestra, serving curated metrics into Tableau and Mixpanel.
Personalisation platform & bespoke CMS: built a content-management and personalisation platform that selects the best experience per visitor in real time using multi-armed bandit algorithms scored against the customer's profile.
AI on AWS Bedrock
The AI features I ship run on AWS Bedrock. The chatbot and the Demystify scanner on this site both call Claude models through Bedrock's Converse API, keeping inference behind a single managed endpoint with IAM, region routing and cost controls handled in the platform.
Stack
Deep: C# 13 / .NET 10, AWS (Lambda, ECS, SQS, EventBridge), distributed and event-driven systems, Snowplow / Snowflake / dbt, AWS Bedrock (Converse API), GitHub Actions and CI/CD (AWS CodePipeline, CodeBuild), code security (SonarQube, CodeQL, GitHub Advanced Security), API and data-contract design, SQL (Postgres, DocumentDB). Working knowledge: Node.js, React, TypeScript, Rust, Terraform / CDK, Microsoft.Extensions.AI, Semantic Kernel, AI agents and skills, custom AI agents, AWS Strands, OpenTelemetry, Tableau, Mixpanel.
Education
Master's in Computer Science (2:1, honours), University of Nottingham, 2012. In industry shipping production systems every year since.
Contact: matt.van.bird@proton.me