Features are not pages
How to derive the screens a product actually needs from its features, and write the flows that expose what is missing.
………0reading9 min readRead field noteBuilding in public
Solo founder · Product systems · AI
I turn product chaos into a clear build plan.
I’m Hovhannes. I help solo SaaS founders decide what belongs in v1, what can wait, and what each decision affects — hands-on in a Clarity Sprint, and with Product Clarity, the planning system I’m building in public.
Product systems · AI-dev · Yerevan
01 / About
Make the system clear before building it.
I turn messy product ideas into clear build plans.
Before the build starts, I pin down the audience, the smallest release worth shipping, and the decisions that affect everything else. You get one plan that both you and your AI tools can follow without guessing.
02 / Current project
In developmentProduct Clarity
A planning system for solo SaaS founders building with AI. It keeps positioning, MVP, pages, flows, and technical decisions connected in one place — so when one feature changes, you see what else breaks before the code does, and your AI tools get the product context, not just a ticket.
Positioning→MVP→Features→Pages→User flows→Roadmap→Dependencies→Technical structure
03 / My principles
Four rules I keep coming back to.
They keep the roadmap honest when a new idea sounds more exciting than the problem in front of me.
- 01
Clarity before code.
- 02
Build less. Make every part earn its place.
- 03
Give AI the product context, not just a ticket.
- 04
Every feature should solve a problem someone actually has.
04 / Yerevan Product Coffee
Six builders. One table. Real product conversations.
A curated coffee roundtable for people actively building digital products in Yerevan. Bring one real question, contribute what you know, and leave with a clearer next move.
- 6 people totalHovhannes + 5 guests
- One focused themeNo talks or pitches
- Table forming nowManual selection
Leave with clarity.
05 / Building in public
What I’m learning as I build.
Honest notes on scope, product decisions, and the parts of AI-assisted building that get messy in practice.
A plan is not finished when the build starts
What to do with the plan once v1 ships — closing open decisions, revisiting cuts with real evidence, and what to leave alone.
………0reading8 min readRead field noteFrom an idea to a build-ready plan
The seven decisions that turn an idea into a build order, in the sequence that keeps each one from breaking the next.
………0reading13 min readRead field noteValidation is not a survey
Why asking people if they would use it predicts nothing, and what evidence to look for instead before writing code.
………0reading7 min readRead field note
What a build-ready plan actually looks like
A complete product blueprint, section by section, filled in for a sample product — the format you can copy for your own.
………0reading13 min readRead field note
What an AI agent needs before it writes code
The difference between a one-off task brief and the durable product context an agent should read before every task.
………0reading8 min readRead field noteWrite the constraint, not the instruction
Why agents follow instructions literally and break unstated rules, and how to write requirements as constraints instead.
………0reading7 min readRead field noteA backlog is not a build plan
How to turn a growing inventory of ideas into one believable week of product progress.
………0reading7 min readRead field note
What my dog taught me about building consistently
Why a dependable rhythm matters more than waiting for another perfect day to build.
………0reading6 min readRead field noteWhy more features create less clarity
How a sensible feature list turns into a product nobody can explain, and what to cut first.
………0reading6 min readRead field noteMVP first: what actually deserves to ship
A practical way to protect the one outcome your first release has to deliver.
………0reading7 min readRead field noteBuild logs beat perfect roadmaps
A short record of what actually happened is often more useful than a polished plan.
………0reading7 min readRead field noteSee what breaks before you build
Why a small request can break five other things, especially when an AI agent only sees the ticket.
………0reading8 min readRead field noteWhen an AI agent should stop and ask
A practical boundary between useful autonomy and expensive guessing.
………0reading7 min readRead field note06 / Contact
Need a clearer next move?
Bring me the messy version. In one focused sprint — five to seven days — you get a build-ready MVP blueprint for one product idea: what ships, what waits, and what each decision affects.
$300 founder rate First three founders · then $600
Honest guarantee: if the kickoff call shows I can’t add real clarity, you get a full refund. Full process, pricing & FAQ on the sprint page.