FOULKES / WORK & IDEAS
Make room for bigger ideas.
Test them in smaller steps.
These are directions I’m exploring, not products available to purchase. Each starts with a question that can be tested before committing to a larger platform.
Construction knowledge, made searchable.
Exploratory concept. Historical estimates could become a structured reference library instead of disconnected spreadsheets. The first question is whether a representative sample can be normalized accurately while retaining its assumptions and source files.
A useful pilot would organize project attributes, scope, dates, units, and cost categories. Estimators would review comparable records and missing context. Forecasting or generating estimates would come later, with qualified review and appropriate current data.
First proof: can a reviewer find relevant past work and verify every extracted value?
A workspace for institutional knowledge.
Exploratory concept. Bring selected procedures, project records, and decisions into a searchable workspace with clear ownership and source references.
The hard question is not just whether an assistant can answer. It is whether the right person can see the right information, recognize an outdated source, and get help when the evidence is incomplete.
First proof: evaluate a small, permissioned collection against real questions and deliberately unanswerable ones.
An assistant inside the process.
Exploratory concept. An assistant could prepare a structured request, identify missing details, and propose a next action inside an existing business workflow.
Start with preparation and review. Only add permission to change records after the boundaries, approval rules, and recovery behavior are clear.
First proof: does the proposal help a person make a better decision without hiding uncertainty?
What’s the problem
you want to solve?
Tell me what you’re working around today. I’ll help you explore the options and be straightforward about the skills, information, and next steps involved.
A shared camera as context.
Exploratory concept. With a person’s explicit permission, a camera-sharing application could let an assistant discuss what they are showing: an object, a workspace, or a step in a process.
The first experiment would test useful visual context, clear start and stop controls, and what happens when the view is ambiguous. Privacy, retention, and access would be defined before use. This is an idea to explore, not an available product or a substitute for qualified safety advice.