A pile of side projects on the home lab: RAG, knowledge tooling, small internal helpers.
Ivan Mišić · Product leadership · Tech · AI
I SHIP product
for a living.
Twenty years of receipts. Field notes on product leadership, digital transformation, and building things that sometimes work.
Latest writing
all 46 →
See which page ranks for each query with the Search Console API.
Search Console exports separate queries from pages, hiding which URL Google shows for each term. The API joins them. Here is the fifteen-minute setup.
A price is simple until the customer qualifies for it.
A guest can start with one offer, then qualify for another when the customer context becomes known. Commercial must define that change, and every touchpoint must keep the price true.
A checkout countdown has to earn its delivery promise.
A checkout countdown can create urgency or support a real delivery promise. That promise needs clear rules for what changes at zero, which date to show, and how it survives payment.
Codex vs Claude Code: two agents on one codebase
I run Claude Code and Codex daily, sometimes on the same repository. My session logs show the context cost, while the working tree shows the coordination cost.
The person logging in is not always the customer.
Login proves who is at the screen. It does not prove what they own, may manage, or can buy. Model those questions separately before the edge cases arrive.
What using Claude Code actually looks like
Using Claude Code is a loop: ask, change, check, decide. Here is what a normal session looks like, six commands worth learning first, and how to recover when it goes wrong.
Capturing email and calendar without a mailbox API
Microsoft Graph and IMAP were off the table, so I used Power Automate, OneDrive sync, and one 15-minute scheduled task to pull email and calendar into a local inbox.
Most of my AI pipeline isn't AI
By volume, most of my AI pipeline is plain Python. The model is a small, expensive part I call as little as possible. The script-first split, and why it wins.
Claude Code skills: building a workflow that runs itself
The control layer of my work system is markdown and small scripts. How a workflow combines Claude Code skills, rules, model routing, and parallel sub-agents.
The Claude Code plugins I built and run
I put the Claude Code tools I reuse across projects into one public toolshed: generators, workflow commands, anti-slop checks, search visibility, rule packs, and prompts.
My AI was smart. It just couldn't see my job.
My work context was scattered across email, meetings, documents, and notes. So I wired it into a private Obsidian vault Claude Code can read.
Your next team is three people and a pile of agents.
I expect AI-assisted teams to become smaller and organize around complete workflows. The model works only when the pod owns the outcome and runs what it builds.
Four calendar days, one person, one full redesign
A full site redesign in four calendar days, from design system to admin stack. The real question is what product teams should do with the time AI gives back.
Klarna's AI U-turn isn't a failure story
Klarna reported faster support and stable satisfaction, while its CEO later said its customer-service cost-cutting went too far. Product leaders need to read both.
The permission tax: when growing teams ship less, not more
It took a month to build and three to four months to get permission. Team growth can slow delivery when every decision needs another approval.
Multi-market apps: the central vs. local paradox
How to run a multi-market app with one shared codebase, local deployments, clear decision rights, and room for real market differences.
AI made building easy. Knowing what to build just got expensive.
AI can make execution cheaper, but it also exposes weak decisions faster. The valuable skill is defining work clearly enough to test whether it matters.
Digital transformation reality check
The programmes I have seen stall did not stall on technology. They stalled on who was allowed to say no, on roadmaps nobody had pruned, and on systems everyone agreed to replace and nobody could afford to. Eight pieces on what actually moves: how a digital team has to be set up before it can ship anything, how to decide what is worth digitalising at all, and what to do about the systems you are stuck with. Each one comes out of work I did, not out of a framework. Read it front to back for the argument, or take any piece on its own.
How to use Claude Code
Claude Code is Claude working inside a folder on your computer. It reads your files, edits them, and runs commands, and you tell it what you want in plain English. No code to write, and no development background needed to start. Twenty-four short chapters, in five modules, take you from opening a terminal for the first time to a website that is actually live. After that, the habits that keep it useful once the project outgrows a handful of files. Start with the overview, which is the whole loop on one page. From there the chapters stand on their own, so you can read in order or open the part you need.
Take something useful
Open Toolshed →
09
Skills
SKILL.md packages Claude loads for specialized tasks.
06
Plugins
Bundles of skills, agents, hooks, and MCP servers.
05
Rules
Project rules and standards Claude follows.
46
Prompts
Copy-ready prompts for specific jobs.
On my board · 39 tools
The AI and product tools actually on my board, with an honest take on each.
Open the board
Current focus
full /now →
tracking
What telcos sell once connectivity isn't the product anymore.
Open current focus
01 · work
02 · tools
03 · thinking
Claude Code in meetings, in the day job, in side projects. Pretty much everywhere.
The gap between knowing-what-to-build and being-able-to-build.