Insights
Practical writing for founders and company owners shipping AI.
Notes from 50+ products shipped across AI-native software, regulated industries, hospitality, automotive, retail, and construction. Written for the founder making real budget calls, not for the conference keynote.
Strategy
When to build AI, when to wait, and how to size the bet.
8 min read
Does your startup actually need AI? A founder's decision framework
Most startups that say they need AI do not. A practical decision framework for founders, with the four signals that mean you should ship AI now and the three that mean you should wait.
Read
9 min read
The five AI projects that almost always fail and why
Five AI project patterns that consistently burn budget and produce nothing usable, and the structural fixes that make each one work. Drawn from 50+ products shipped.
Read
9 min read
Build vs buy AI: when to use foundation models and when to train your own
When does it make sense to call OpenAI, Anthropic, or open-weights, and when do you actually need to train? A practical decision tree for non-technical founders.
Read
9 min read
What an AI MVP actually costs in 2026
A real cost breakdown for AI MVPs in 2026, from a 30,000 euro narrow workflow to a 200,000 euro multi-tenant platform. Line items, timelines, and the cost drivers founders miss.
Read
Methodology
How real AI products get shipped without burning the budget.
10 min read
The 90-day AI MVP plan for non-technical founders
A week-by-week 90-day plan for non-technical founders shipping an AI MVP. The artifacts, the meetings, the milestones, and the four checkpoints where most projects die.
Read
9 min read
How to write an AI product spec your engineers will actually build
A template and worked example for writing an AI product spec that survives engineering review. The seven sections, the failure modes, and the line that decides whether the project ships.
Read
9 min read
Why most AI proofs of concept never reach production
AI PoCs that look impressive in a sandbox die at the production gate. The seven structural reasons why and the rebuild strategy that gets a stalled PoC across the line.
Read
10 min read
RAG, fine-tuning, or agents: a founder's guide
When to reach for retrieval augmented generation, when to fine-tune, and when an agent is actually the right answer. Cost, latency, and risk trade-offs in plain language.
Read
9 min read
The hidden costs of AI features: tokens, latency, evals, hallucinations
Token cost is the line founders see. The line items they miss are latency engineering, eval upkeep, hallucination handling, and on-call. A real cost picture for AI products.
Read
Verticals
Where AI moves the needle in restaurants, dealerships, construction, retail, and compliance.
9 min read
AI for restaurants and hospitality groups: where it works and where it does not
Practical AI use cases for restaurant groups and hospitality operators in 2026. The four workflows where AI saves money this quarter and the three pitches that waste it.
Read
9 min read
AI for car dealerships: what 20+ rooftops taught me about what actually works
Lessons from AutoPlatform across 20+ dealerships in Italy, Switzerland, and Albania. The AI workflows that grow used car gross, the ones that fizzle, and the integration realities.
Read
9 min read
AI for construction companies: from proposal automation to risk forecasting
Lessons from ConstructionOS shipping AI proposal automation. Where AI cuts six figures of senior estimator time, and the workflows construction operators should leave alone.
Read
9 min read
AI for retail loyalty programs: lessons from reaching 250,000+ families
Lessons from a multi-tenant loyalty platform reaching 250,000+ families across major Albanian retailers. The AI workflows that lift basket size, redemption, and retention.
Read
9 min read
AI for regulatory compliance and audit: lessons from RegNexa
AI workflows that compress compliance and audit cycles in regulated industries. The four use cases that work, the trust patterns that hold up, and the failure modes to avoid.
Read
Vision
What changed in the last 18 months and what it means for company owners.
9 min read
The strategist-builder: why one senior AI PM can replace a five-person team
The collapse of the gap between strategy and execution is the most underappreciated shift in product in 2026. Why one senior AI product person can now do the work of a small team.
Read
9 min read
How AI tooling collapsed the cost of shipping software
The cost of shipping a real software product fell by an order of magnitude in 18 months. The numbers, the mechanism, and what it means for company owners planning their next product bet.
Read
Working on something where AI strategy and hands-on execution both matter? Email is the right first step.