This archive provides insights and practical notes on various aspects of enterprise AI, helping professionals navigate the complexities of AI product strategy and implementation.
Every entry is one drop — a short, evidence-first field note from building AI in production. Together they span agent engineering (when to use an agent, how to evaluate and ship one), model economics (what an LLM feature really costs, build versus buy), evaluation and reliability, security and permissions, and enterprise adoption. Browse by topic or by month below, or start with the newest drop. Last updated September 2026.
AndyMental is one free, evidence-first drop a day on applied enterprise AI — agent engineering, model economics, evaluation, security and adoption — written by a practitioner shipping it in production.
# What you get
Price
Free — no paywall, no gated archive
Cadence
One new drop every day
Archive
357+ drops, all readable in full
Formats
Field-note essays, TILs with working code, quick takes, curated links and videos
Newsletter
Optional and free — one drop a day by email, one-click unsubscribe
Author
Anand “Andy” Padia, Global Head of GenAI, Trigent Software
# Who writes this
Anand “Andy” Padia — Global Head of GenAI at Trigent Software, author of Orchestrating Intelligence, building AndyMental in public.
“Share without worrying. Ship daily. Change by sharing.”
# Why it’s different
Evidence-first: every claim names a specific model, paper, incident or launch with a date, or is labelled as judgment.
Written by a practitioner shipping enterprise AI in production, not a commentator summarising the news.
Contrarian when the evidence allows, never hype and never doom — one original claim per drop.
Free, daily, and readable in full — no paywall, no gated archive.
Daily, evidence-first field notes on applied enterprise AI — agent engineering, model economics, evaluation, security and enterprise adoption — written by Anand “Andy” Padia, Global Head of GenAI at Trigent, from building AI in production. One drop a day.
Is it free?
Yes. Every drop is free to read in full, with no paywall and no gated archive. The optional email newsletter is free too.
How often is it updated?
One new drop every day. The archive currently holds 357+ drops and grows daily.
How do I follow along?
Subscribe for one drop a day by email, follow Andy on LinkedIn, or use the RSS feed. If you are new, the Start here page lays out a path through the six main threads.
Who is it for?
Working engineers and tech leads shipping AI in real organisations — readers who want evidence and a decision they can act on, not hype.
Where do the claims come from?
Every drop names its source — a specific model, paper, incident or launch with a date — and links the primary reference. First-person practitioner judgment is marked as such rather than dressed up as fact.
