Skip to content
View harmenx's full-sized avatar
🤠
Focusing
🤠
Focusing

Block or report harmenx

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
harmenx/README.md

Hi, I'm Harm 👋

AI Engineer · Full-Stack Engineer
I ship LLM products and high-availability systems to production: Bedrock, RAG, React and AWS.
📍 London, UK

Website LinkedIn 6x AWS Certified


⚡ At a glance

  • 14+ years building production software for startups, enterprise and government in the UK and New Zealand
  • Built Xero Spotlight, an AI analytics product on Amazon Bedrock + RAG, designed for 100 messages/second
  • Rebuilt navigation for 4.5M customers with zero downtime on a tier-zero service
  • Delivered the first production LangChain system on AWS at Joker Seven
  • Conducted 600+ technical interviews for Amazon, Pinterest and Wayfair via Karat

🧰 Tech stack

AI & LLM
Amazon Bedrock LangChain LangGraph SageMaker · RAG pipelines · prompt engineering · vector databases & embeddings · agentic workflows · LLM evaluation

Languages
Python TypeScript JavaScript Java C# Swift

Frontend
React Next.js React Native Electron D3.js · WebGL · micro-frontends · SSR · WCAG 2.1 AA

Backend & Data
Node.js .NET Kafka GraphQL PostgreSQL Redis DynamoDB MySQL Snowflake · microservices · REST

Cloud & DevOps
AWS Docker Kubernetes Terraform Jenkins Datadog New Relic · Lambda · S3 · EC2 · SQS · API Gateway · KMS · VPC · boto3 · CI/CD · SLOs


📌 Selected work

Delivered for employers and clients, so the code is private.

Project What I did Impact
Xero Spotlight AI analytics that generates LLM messages from historical customer data. Kafka + .NET consumers on Bedrock with a RAG pipeline and LLM output evaluation against historical and synthetic data. Built for 100 msg/s
Xero Navigation Rebuilt the primary navigation of a tier-zero service in React + TypeScript, with feature flags and redundancy layers. 4.5M customers, zero downtime, 99.999% uptime
Xero Platform Rebuilt the localisation architecture, set component and WCAG 2.1 AA standards, and moved SLO provisioning to Terraform. 50% fewer deployments, 20% fewer UI defects
Joker Seven AI Identified AI use cases and shipped the company's first production AI system (Python, LangChain, AWS), plus a SageMaker Pipelines project. First AI system in production
Amazon Merchant Portal Rebuilt the SCP UI and designed the Java workflow API layer. Saved ~3 months of rework, shipped early
Payment Pipeline High-throughput payments across Checkout.com, PayPal and Adyen, with 3D Secure, refunds and mobile payments. DynamoDB processing 80% faster
Archax Trading FDC3-compliant institutional trading dashboards (Electron, WebSockets, D3.js) with zero-downtime auto-upgrades. Sub-10 ms latency
Twig Education Micro-frontend architecture that removed cross-team deployment coupling. 40% less release overhead
Rackspace Logistics Rebuilt an on-premise logistics app as a cloud-native service. 50% lower latency
NZTE Export Portal Government export portal integrated with Okta and Dynamics. 5,000+ businesses, 80% faster loads, ~70 NPS

🛠️ On GitHub


🏢 Career
Role Company When
Lead Trainer & Assessor Institute of Data, Auckland: full-stack, cloud and AI bootcamp cohorts, 100% pass rate 2026
Senior Software Engineer Xero, Auckland 2025–2026
Technical Lead Joker Seven, Auckland 2025
Senior Software Engineer Amazon, London 2024
Technical Lead Machina Consulting, London 2023–2024
Technical Lead Archax, London 2022–2023
Technical Lead Machina Hero, Edinburgh: Nucleus Financial, Twig Education, Zonal, VisitScotland, Rackspace 2019–2022
Technical Lead NZTE, Auckland 2018–2019
Technical Interviewer Karat: 600+ interviews for Amazon, Pinterest and Wayfair 2019–2022
Earlier NZ Transport Agency · Propellerhead (AT Mobile, 200k+ users) · DeRef · Solution Management · Money Mappa 2012–2018

Education: BSc Computer Science & Information Systems, University of Auckland

🏅 Certifications & awards

AWS (6x)

  • Solutions Architect – Professional (SAP-C02), 2023
  • Machine Learning Engineer – Associate (MLA-C01), 2026
  • Security – Specialty (SCS-C02), 2026
  • Advanced Networking – Specialty (ANS-C01), 2026
  • Developer – Associate (DVA-C02), 2026
  • Solutions Architect – Associate (SAA-C03), 2021

Awards

  • 1st Place, NZTA Hackathon (2018)

Systems, not goals. Build once, compound forever.

Pinned Loading

  1. awesome-faang awesome-faang Public

    Resources for preparing for FAANG and big tech coding, system design, and behavioral interviews.

    4 4

  2. awesome-passive-income awesome-passive-income Public

    Income that continues after the work that created it is finished.

    7 2

  3. awesome-startup awesome-startup Public

    Tools for taking a startup from idea to launch and scale.

    1 3

  4. adhd-dev-survival-guide adhd-dev-survival-guide Public

    ADHD-Optimized Developer Toolkit

    1

  5. aegis aegis Public

    A real-time lobby: a night-earth globe, activity bars, and a live event stream.

    C#

  6. venice-mcp-server venice-mcp-server Public

    Forked from veniceai/venice-mcp-server

    Venice AI MCP Server for Claude Code, Cline, and AI agents

    TypeScript