Suggestions
Julian Harris
AI and climate tech, specialising in how to design and build high performance AI systems.
Julian Harris is a technical product leader with over 25 years of experience in creating internet software that people love, including 7 years spent at Google. Passionate about combating climate change, Julian left his job to focus on using his skills to make a positive impact. He has specialized in the impact of deep learning, including NLP, on product design since 2018 and is currently working on a new venture to promote happier and more sustainable lifestyles.
Throughout his career, Julian has co-founded three companies, two focused on internet technology consulting and one integrating social media into museums and galleries. He has held technical leadership positions since 2003, overseeing teams, programs, and products with global reach and has consistently stayed close to evolving technologies.
During his time at Google, Julian primarily served as a product manager, guiding product delivery on a global scale and designing internal sales products to enhance Google's advertising sales operations. He led engineering teams, employed agile and scrum methodologies since 2006, and maintained stakeholder networks across the globe.
Before Google, Julian worked on innovative projects in the UK, Barcelona, the Bay Area, and New Zealand, focusing on agile project delivery in Java. Notably, he was the tech lead for projects like Friends Provident eSelect and Telefonica's 'PDSL' project, recognized for their innovation and impact.
Julian's educational background includes studying Computer Science at The University of Auckland. Additionally, he has been involved in various organizations and roles, ranging from founder and CTO at Knowcast to advisory positions in companies like Inicio AI and EARTHLY. Julian's diverse experience also includes working as a Senior Product Manager at Google, serving in technical roles at Department for Work and Pensions (DWP), and engaging in emerging technology consultancy.
Passionate about climate change, technology, product design, deep learning, agile methodologies, and sustainable living, Julian Harris continues to make a significant impact in the tech industry and beyond.
Highlights
AI coding agents are TERRIBLE at finishing jobs.
Refactoring is the perfect storm:
- happens all the time
- code-base-wide changes
- very very hard to get it fully complete.
You HAVE to have multiple iterations to catch stragglers. I had one project which needed FOURTEEN iterations to make.
(this is built-in to Ceetrix, my latest venture.)
#ceetrix -- MCP-based AI Delivery Manager Agent

Remarkable. Live demo seems to be down currently though.
