Site AI
Ask shiran.io
A site-specific assistant designed to help visitors find the right article, understand a concept, or continue from something already published here.
Knowledge → question → useful answer
AI Lab
Useful AI, tested against real work.
I use this space to test where AI can genuinely improve research, trade, operations and decision-making — and to separate useful capability from impressive demos.
Featured experiment
Try something useful.
The Lab should give visitors something to do, not just something to read.
Experiments
What I’m building and testing.
Small products and practical ideas that use AI for a clear job.
FreightIn development
Freight quote explainer
Explain quote lines in plain language, surface likely exclusions and make comparisons easier to understand.
View experiment→
BusinessExploring
Process improver
Break down a repetitive workflow and identify where simplification, automation or AI could remove friction.
View experiment→
TradeExploring
Supplier question builder
Turn a sourcing requirement into clearer supplier questions before price becomes the only discussion.
View experiment→
DocumentsPlanned
Trade document explainer
Help a reader understand the purpose and structure of a freight or trade document while keeping human judgment visible.
View experiment→
BooksExploring
Book → action
Turn a useful idea from a book into a practical action, checklist or decision framework.
View experiment→
SiteOngoing
Ask shiran.io
Use the site’s own material as context for finding, understanding and applying what is already here.
View experiment→
What I’m testing
Products worth paying attention to.
Not a giant “top AI tools” list. Just products I’m actually exploring and what I’m learning from them.
01
Research & reasoningWhere deeper research support genuinely saves time, and where source-checking still matters.
02AI inside work toolsWhether AI is more useful when it lives inside the workflow instead of becoming another separate app.
03Computer-use agentsWhat changes when AI can carry out multi-step work rather than only explain what to do.
04Focused AI productsSmall products that solve one specific job well instead of trying to be everything.
Lab rules
How I decide what belongs here.
The Lab should stay useful even when AI trends change quickly.
01
Clear job first
AI should have a specific role. If the user cannot understand why it is there, it probably should not be there.
02
Show limitations
Useful experiments should make uncertainty and human judgment visible instead of pretending AI is always right.
03
Keep what earns its place
An experiment stays only if it saves time, improves clarity, helps a decision or makes something meaningfully easier.
More than experiments
Use the rest of the site too.
AI is one part of shiran.io. Articles, tools and practical resources remain the foundation.