Harry Johnson

Product Engineer at Eigen Technologies

harry ]dot[ johnson500 ]at[ `gmail` ]dot[ com

[CV]

About

Experienced Software and Product Engineer currently working at AI scale-up Eigen Technologies (opens new window). Co-leader of the MLOps team, we build data and compute pipelines to productionise research models for unstructured-data extraction.

My background is in Python for data analytics and process automation, where I found familiarity with the PyData stack, and did some early experiments with Data Science techniques. I used statistical and qualitative approaches to assess product strategy and competition at a 10K employee software company, where I implemented internal software projects, and delivered reports to the C-Suite. I gained expertise in Domain Driven Design (DDD) coding against a large in-house ERP system in PHP, where I learned the discipline of Software Engineering and the intricacies of client relations.

In my spare time, I am learning Rust and practicing Data Science, with an interest in parallel processing.

Experience

• Product Engineer
  Eigen Technologies: 2021 - present
  Co-Leading the ML-Infrastructure team: I am responsible for working with Data Science, Product Management, and the ML-Infrastructure team to productionise our machine learning capabilities. I manage the migration of our legacy embedded ML infrastructure to a Kubernetes-first microservice configuration, as well as the implementation of new features. I lead design sessions, co-ordinate with stakeholders, manage timelines, and contribute to implementation. I prioritise scalability, modularity, extensibility, and lowering the long-term maintenance burden with rigorous attention to software interfaces and data contracts.
  Before co-leading ML-Infrastructure: Contributed to design and implementation of a microservice-oriented DDD refactoring of critical project models, collaborated developing an async client for remote system testing, tools for system observability on client deployments, and various reporting features.

  • AWS (EC2, S3, ELK)
  • Azure
  • Dagster
  • MLFlow
  • BentoML
  • Kubernetes
  • Helm
  • Docker (+Compose)
  • Python 3
  • PostgreSQL
  • Redis
  • Celery
  • Microservice Architecture
  • React
    
    

• Software Engineer
  Gigabyte Software: 2018 - 2021
  Full-stack development - I was responsible for liaising with clients to collate requirements, designing and implementing new functionality in our proprietary ERP system, while also maintaining the existing codebase. On occasion, I was responsible for resolving infrastructure issues. I managed and worked on the Barclays Bank integration project, liaising with Barclays and our client, and planning and executing the design-implementation-test cycle. I designed and implemented (with the team) a supplier management module, and a postal mail integration with ClickSend, among others.

  • PHP (5.6, 7.4), Symfony, Doctrine
  • MySQL
  • Ubuntu Server
  • Redis
  • Beanstalk
  • DDD
  • React (+Redux, +Saga)
  • Heroku
  • AWS S3
  • AWS CloudFront
    
    

• Competitive Intelligence Analyst, Product Strategy
  Finastra, London: 2016 - 2018
  I managed the implementation of a 3rd party innovation application, and designed and managed the implementation of a custom post-deal survey component in the company Salesforce instance. I used pandas and the PyData stack to produce quarterly reports for the C-Level. I built a Random Forest Classifier to predict deal outcomes, and Flask web application to generate reports. I used the Python web-scraping tools to collect data on competitors, and pandas and PyData to combine it with deal information to produce heatmaps for regional sales targeting.

  • Python (2, 3)
  • Flask
  • pandas, numpy, Plotly, Bokeh
  • pptx
  • Salesforce API
    
    

• Business Intelligence Analyst, Sales Operations
  Misys (now Finastra), London: 2014 - 2016
  Working in Sales Operations, I identified multiple manual processes available for automation. I automated various operational and reporting processes I was tasked with, including the Global Deal Review and Monthly Sales Forecast, using the Python Win32COM APIs, MS Office Python packages, Salesforce integration, pandas, and a variety of visualisation tools.

  • Python (2)
  • pandas, numpy, Plotly
  • Salesforce API
  • pywin32
    
    

• [Please see my LinkedIn (opens new window) for further employment history]

Expertise in Technologies and Methodologies

• Domain Driven Design

• Software and Data Architecture

• Agile

• Microservice Architecture
  
  

• Python

  • Dagster
  • BentoML
  • SKLearn
  • HuggingFace
  • Django
  • Flask
  • Pandas (+ PyData ecosystem)
  • Wagtail CMS
  • Django Rest Framework
  • Scrapy (BeautifulSoup, Pattern, nltk)
  • Celery
    
    

• Devops

  • Docker
  • Vagrant
  • Ansible
  • Kubernetes
    
    

• Cloud Providers

  • AWS
  • Azure
    
    

• SQL

  • MySQL
  • PostgreSQL
    
    

• NoSQL

  • MongoDB
    
    

• PHP

  • Symfony
  • Doctrine
    
    

• Javascript

  • VanillaJS
  • ES6
  • React (+Redux, +Saga)
    
    

• Linux Server Administration

  • Bash/Shell Scripting
  • Apache Web Server
  • Logging
    
    

Education

• University of Edinburgh
  2009 - 2013

• Mander Portman Woodward College
  2007 - 2008

• Eton College
  2003 - 2007