Cultivating useful tools and calm workflows with AI.
A quiet space to learn how to guide modern language models in your everyday life, automate tedious tasks, and build software that genuinely serves people.
No speculative hype, no jargon—just hands-on lessons, working applications you can test, and honest observations from a builder.
The Philosophy
The story behind "Good Soil"
The name comes from the familiar parable of the sower. When seeds are scattered across a field, much of what happens next is simply outside the planter's control. Some fall on hard paths and birds sweep them away. Some fall on shallow rock and wither in the heat. Others are choked by thorns and weeds.
Building software and adopting new technology feels very similar. Platform changes happen, models hallucinate, projects fail to find traction, and unexpected obstacles get in the way. It is easy to look at the birds, the rocks, and the choking weeds—everything outside our control—and let fear freeze our hands.
Yet if we let fear keep us from scattering seeds, we never get the chance to find good soil.
This site exists to help you plant with confidence: learning how to use AI practically, accepting that not every experiment will be perfect, and continuing to build until you produce something genuinely helpful and lasting.
Workflows
Using AI in daily workflows
Step-by-step guidance on weaving language models into your daily routines—saving time on tedious tasks without losing human oversight.
Guide 5
Letting AI do the boring parts
You'll be able to pick one repetitive task and design a safe routine for it.
Guide 4
Asking better questions
You'll get usable answers on the first try more often.
Guide 1
What today's AI actually is
You'll be able to say what today's AI is, in one sentence, without using the word AI.
Guide 2
How a chatbot builds an answer
You'll understand why a chatbot can be fluent and wrong at the same time.
The Workshop
Tools built for real life
You don't need a corporate software team to build helpful things. Using coding agents with careful human oversight, I build small web applications that solve specific daily needs.
From idea to working tool
Each application below was planted to meet a concrete need for students, church groups, musicians, or community events. Every line of code was generated with AI coding assistants and reviewed line by line before deployment.
See all 10 working apps
Good News
No account neededA guided, visual way to share the good news of Jesus in under seven minutes.
Read overview→Professor Turtle
Free, sign in with GoogleAn AI study companion for Singapore students and parents.
Read overview→Quizzard
No account neededRealtime trivia, conjured for your crowd.
Read overview→Lifted
Passcode from the hostCell Group Prayer & Praise Wall.
Read overview→About the Maker
Hi, I'm James.
I'm a middle-aged technology manager based in Singapore. During the day, I manage enterprise technology and software projects. In the evenings, I build tools and write about technology on my own.
I don't sell courses, run advertising, or promote AI as magic. I believe in using modern models with careful judgment—treating them as powerful statistical tools that require thoughtful oversight, patience, and good soil.
Field Notes
Latest writing
Practical essays on software architecture, checking model outputs, and where AI still slows me down.
Judgement
How I check what AI tells me
A practical framework for evaluating machine-generated text, identifying plausible hallucinations, and knowing when verification is mandatory.
Building
Nine small apps, and what they taught me
Reflections on building and deploying nine small web applications, from public tools to software built for small groups.
Using AI
The approval gate is the feature
Why automated workflows must retain a deliberate human checkpoint whenever systems interact with colleagues, clients, or production systems.
Planting good seeds
Technology comes and goes, but clear thinking and useful work endure. Explore the guides, test the tools, or reach out.