Hunger on Tinkering
Long ago, we thought the magic was in the separation.
The Developer would obsess over cool code structure. The Designer would fixate on aesthetics, chasing colors to make it "pop."
But I felt a silence in the middle. While we were busy bringing ideas to life, We almost forgot how to bring results.
We were building "stuff"—features, pages, updates— But struggling to solve the actual problem for the human on the other side. Value was getting lost in the process.
That’s why my purpose shifted. I am a builder who looks at the whole. Not just writing code. Not just designing views. But optimizing for what matters.
I realized that User Experience is a strategic advantage. The difference between a visitor and a believer. When you combine a heart focused on results with the hands of a maker, The goal changes.
It is no longer about building software. It is about building meaningful products. A future where we stop assembling parts, And start creating impact.
The work now: results over features, production constraints as design material, and judgment as the durable edge.
I'm mosh.
Begin Chat →Focused on creating intuitive and performant web experiences. Bridging the gap between design and development.
Selected Projects
Advanced components and templates to craft beautiful websites.
UI kit to make beautiful, animated interfaces.
Work Experience
Founder
Hawklist
2026
Founder and Product Strategist
ConvFetti
2024
Head of Product Experience
Luetone
2025
Product Designer
Nordstone
2023
Senior Brand Designer, Marketing
What's Goin On Qatar
2022
Senior UI/UX Designer
TeraBlock
2022
Designer, R&D
Qatar University
2021
CPO
Go Laundry
2019
Product Designer
WorkJam
2019
Product Manager
Worldtalentjobs
2018
Web Designer and Developer
Worldtalentjobs
2017
Builder, Co-Founder
Rora Media
2016
Webmaster
Grand Mart Group
2015
User Experience Developer
Fiume Technologies
2013
Blog
SaaS activity tracking: designing an admin log teams trust
Admins needed to see every team action in one place. Research with two managers and eight user stories turned a confusing log into filters, search, share, and export.
GymProLuxe: turning a resistance kit into a training system
GymProLuxe sold trusted hardware but motivation faded after delivery. Premium app UX gave every owner a daily plan matched to their kit.
ScoreAi: match search redesign that lifted engagement
ScoreAi had demand but its match search leaked engagement. A conversion-first rebuild of hierarchy, scanning, and onboarding fixed discovery.
You don't have a prompting problem. You have a vocabulary problem
Ben Tossell shipped a style picker because founders can't name what they want — and Google's DESIGN.md proves a shared vocabulary beats a longer prompt every time.
Your AI stack just shrank to three layers. Cut the rest
Vercel shipped Fluid and five models to one gateway in six days while Claude unified memory across chat and Cowork. The solo stack is now three layers — and the extra tools are just weight.
Figma shipped generative plugins. Your system decides if they're useful
Figma now lets you prompt an agent to build native plugins and animated shaders, with code you can view and ship to React. Solo founders win by holding the system tight, not by generating more.
Prompt engineering is dead. Context engineering is the job
Prompt tricks stopped moving the needle. Context engineering — the files, constraints, and memory your agent sees before it acts — decides if your AI product works twice.
Calm UX is the AI advantage in 2026
Every AI feature adds decisions. The products winning in 2026 do the opposite — they remove them. Here is what calm UX looks like and how solo founders can ship it.
You let the agent decide. The edit is still the job
A solo builder told Claude "you decide" and shipped a design users thought was hired out. The trick was not the prompt — it was the system and the edit after. Here is how to use "you decide" without losing taste.
Your agent can't see your taste until it's readable
Your design taste lives in your head until you write it where agents can read it — Figma variables, Code Connect, and a short taste file. Here's how to make taste machine-readable so every generation stays on brand.
Your prototype works once. Determinism ships it
Vibe coding gets you a working demo in an afternoon. Deterministic engineering — contracts, retries, evals, and rollbacks — is what makes that demo work the hundredth time without you watching.
Google spam policies, fake freshness, and why your dates matter more than you think
We audited our own sites for Google publication-dates compliance and found sitemap lies, back-dated pillars, and silent parseDate fallbacks. What Google actually checks and how we fixed it.
A demo proves nothing. Ship on an eval gate instead
A weekend AI demo hides its failure rate. An eval gate — a quality bar your changes must clear before shipping — is what turns a prototype into a product you can trust at scale.
Building got easy. Distribution didn’t — and that’s the real problem now
AI collapsed the cost of building a product, so the scarce thing in 2026 is a first user who comes back. Distribution, not development, is the bottleneck for solo founders and small teams.
AI startup ideas are compressions of work that already exists
The winning AI products of 2026 do not invent new behavior. They compress repetitive work people already do by hand. Here is what a workflow compression is, why it beats invention, and how to find one.
Your AI product is one model update away from being a free feature
Model providers are absorbing the features their startup customers built first. Getting platformed erases a product in a single provider announcement — unless it owns a workflow, proprietary data, or an interface that is the product. Here is the test.
Agentic AI won’t fix the interface problem
The agentic shift moves AI from suggestion to action, but the moment a human has to take over is still where trust breaks. Here is why autonomy does not solve the interface, and what to design instead.
Everyone is building the same app: a blue gradient on a dark background
An analysis of 210,759 real AI design prompts shows the default look of 2026: a blue gradient on a dark background, with builders referencing Linear more than Apple. Here is why everything converges and how to opt out.
Your AI agent has no memory. Context files are the fix
Every coding session starts from zero. Over 60,000 open-source projects now ship an AGENTS.md so agents stop guessing. Here is what belongs in yours, and why bloated files make agents worse.
The AI build stack collapsed. The leftover work is the product
Three years ago, shipping a production app meant a DevOps team and ten grand a month. In 2026 a solo founder runs one on a few hundred dollars. The stack collapsed into platforms — and the leftovers, the contracts and the eval gate, are the entire product now.
AI is for the plumbing, not the judgment
Anthropic's own survey data says experienced users name judgment as the capability AI still lacks. The solo founders winning in 2026 automate the grunt work and keep every call that needs a point of view. Here is the split that separates them.
The median AI product is getting worse. Early retention is the new moat
Stripe's latest solo founder data shows median startup revenue fell last year while the top 10% grew — a record 61x gap. The differentiator is early retention, and early retention is a design problem.
AI forgets your taste between sessions. Write it down
First-pass AI output is the median of the internet until you give the model a review rubric to fail against. Models forget your taste between chats; a written file keeps it. Here is what a taste rubric checks and where it should live.
Design drift is the tax AI charges on generated UI
A single AI screen looks fine. Twelve of them look anonymous. Design drift is the accumulation of near-misses that costs your product its identity — here is what it is and how solo founders stop it.
Designers are building their own AI tools now. That is the real shift
Half of designers now ship code, and a growing share build their own microtools. For solo founders, that toolmaker instinct matters more than the tools themselves.
Vibe coding raised the floor. Agentic engineering raises the bar. Taste runs the gate
Karpathy named the discipline that follows vibe coding. Agentic engineering pushes AI agents further, and the whole game becomes the review gate. A review gate runs on taste, and a solo founder is the only one holding it.
AI products are non-deterministic. Design like it
An LLM gives you a different answer every time it is called. The product that wins does not hide it — it designs for it: showing limits, letting users steer, and giving failure a way forward.
The last 20% is the whole point
AI nails the first 80% of a product in minutes. The last 20% — hierarchy, edge cases, copy, point of view — is where taste lives, and it is still your job.
Somebody still has to care
AI can write your code, but it cannot care. The human filter — nitpicking, obsessing, fixing — is what makes an app memorable.
AI Design Engineering: How Product Teams Ship Faster Without Losing Quality
A practical guide to LLM workflows, agent-safe design systems, production UI, and human quality gates.
Why I left my job to start my own company
Why I left a comfortable job to build independently — and what I gained and lost.
What I learned from my first year of independent work
Hustle myths, money cycles, loneliness, scope creep, and the golden rule.
How to Export Metadata from MDX for Next.js SEO
Export metadata from MDX so Next.js can drive titles, descriptions, and canonicals.
The design engineering toolstack is multiplying
Designers now use twice as many tools as last year. Here's what that means for solo builders and design engineers.
Stay at the Great Line
A self-reminder about consistency. Showing up every day is the only thing that actually works.
The best AI tools are boring
Claude Code ships Bun rewritten in Rust — a boring, invisible upgrade. Here is why the best AI tools are the ones you do not notice.
Why I stopped chasing AI models
New models drop every week. Qwen 3.8, GPT-5.6 variants, Kimi K3. Here is why the solo builders who win are the ones who pick a stack and ship.
Design taste is the new moat for AI startups
Vibe coding means anyone can ship a prototype in hours. Design taste — knowing what to build and what to leave out — is now the only durable advantage.
AI slop is drowning the web
Merriam-Webster made it word of the year. 52% of new articles are AI-generated. Here is what AI slop is, why it is toxic, and how to avoid making it.
Why AI startups fail at design (and it is not about lack of talent)
Most AI startups fail at design because they treat it as the last step. Here is what actually goes wrong and how solo founders can fix it without hiring a designer.
The vibe-code gap is real
Vibe coding gets you a prototype in hours. Getting that prototype to production takes weeks. Here is why the gap is widening and how design engineers close it.
Taste engineering is the skill after vibe coding
Prompting got us an answer. Context engineering made it relevant. Taste engineering decides whether the answer is actually good.
The hardest part of building with AI is not the building
AI tools let you build anything in days. Validation, distribution, and judgment have not gotten any easier. Here is what solo founders should actually spend their time on.
Design taste infrastructure is the missing AI stack
New tools package design taste into files AI agents can read. Taste infrastructure is the layer every solo founder needs in their stack.
Distribution is the real bottleneck for solo founders
AI made building almost free. Getting users is the hard part now. Here is what actually works for solo founders in 2026.
What taste tools miss about taste
Taste tools capture your preferences in machine-readable files. But the hardest part of building a good product has never been knowing what good looks like — it is holding that standard across every surface, every edge case, every day.
AI has no taste — and that is your advantage
AI-generated products all look the same because models average what they have seen. Solo founders who override the default win on taste — not speed.
Your AI eval set is your taste, made measurable
An eval set is a list of real inputs with the outputs you consider good. It is also the closest thing AI product development has to written taste — and the discipline most solo founders skip.
AI products fail at the interface, not the model
Users rarely abandon an AI product because the model was wrong. They abandon it because the interface hides uncertainty, offers no way to steer, and treats failure like a bug.
Every solo founder is a design engineer now
Solo founders are 63% of new startups, and the best ones are design engineers who never got the title. Here is how to close the last 20% of the work AI cannot.
The AI aesthetic is a costume. Copy the two parts that are real
The AI aesthetic is now a named thing: sparkles, streaming text, thin icons, beige. Two idioms encode real constraints. The other four are a costume. Here is what to copy and what to skip.
I still write code by hand. You should keep a slice of your workflow that way too
GitHub found Copilot makes developers 55% faster. METR found AI made experienced devs 19% slower. Both are right. Here is the hand-build window that keeps your taste sharp.
Speed stopped being impressive. Slowing down is the AI edge now
Everyone with an AI tool ships fast in 2026. The teams and solo founders who win are the ones who slow down on purpose and apply judgment, not the ones who output the most.
Your AI product doesn't need more flair. It needs consistency
AI made design execution free, so every AI-built product is now polished — and anonymous. Users do not trust flair. They trust the products that stay consistent. Here is how solo founders hold the line.
Adaptive AI interfaces are the 2026 trap nobody names
Adaptive AI interfaces that rewrite themselves per user are the year’s biggest design trend. Without a human deciding what the product should feel like, they build no trust. Here is the trap and the fix.
The open-weight wave is good enough now, and that changes the product math
The August 2026 open-weight wave — Qwen3.8-Max, DeepSeek V4-Pro, GLM 5.3, and 7B models that rival last year’s 70B — did not just move benchmarks. It collapsed the model moat. Here is what stays defensible.
AI products fail when they try to replace instead of compress work
The AI products that die all make the same promise: we will replace this job. The ones that win promise something smaller and truer — we will compress the part of it you hate. Here is why replacement fails and compression ships.
Agentic AI doesn’t need a better prompt. It needs a permission interface
Agents now open PRs and run multi-step tasks on your files. The bottleneck isn’t capability — it’s the missing interface for permissions, approvals, and audit that makes autonomy trustworthy.
The first AI draft is always slop. Here’s the taste check that fixes it
The first AI draft always looks fine until you ship it. A simple 4-check taste review — intent, consistency, hierarchy, and edge states — catches the slop before users do.
You don't need another AI tool. You need fewer
Solo founders who ship fastest in 2026 run 6 to 8 AI tools, not 15. Here is why a tighter stack wins, what belongs in it, and the one rule for cutting the rest.
Why curation is the job now
AI makes generation cheap — a solo founder in Berlin kept 116 of 400 AI concepts, a 29% hit rate. Shipping in 2026 is not about prompting more. It is about picking one and cutting the rest.
Your Figma file just became an agent's workspace
Figma opened the canvas to AI agents through the MCP server and skills. The design file is now where agents read your taste, apply your system, and keep work tied to production code.
When AI remembers, your product decides what to forget
Claude shipped memory that sticks across chat and Cowork this week. That makes remembering easy — and makes deciding what to forget the actual product work for solo founders.
Your AI prototype passes the demo. The second session decides if it ships
Your AI prototype looks finished in the demo. It breaks on the second session — empty states, stale memory, and no way back. Here is the test solo founders use to ship beyond the happy path.
AI stopped being a destination. Now it lives inside the work
Standalone chat was the first AI interface. In 2026 the winners are embedded — AI that lives inside the doc, the canvas, and the inbox where work already happens. Here is why the shift matters and how solo founders should design for it.
Vercel ran it 200 times to make taste repeatable
Vercel turned 200+ agent runs into a three-layer system — a taste file, a stylesheet, and deterministic checks. The result was 57% fewer repeat failures, but every page still needed a human to ship.
Task-specific agents ship. General agents demo
62% of companies now experiment with AI agents but only 23% scale them. The gap is not model power — it is scope. Solo founders who ship pick one workflow and compress it instead of promising everything.
OUI-1 proves generative UI lives or dies on format, not model size
OpenUI shipped OUI-1 at 71.7% on the Generative UI Benchmark with 4B active parameters. The win was not bigger weights — it was a line-oriented format, a parser as reward, and self-distillation that cut tokens and errors at once.
Taste is judgment, not preference — and AI can’t fake it
AI can generate anything in seconds, but it can’t decide what’s worth shipping. Taste is judgment under constraints — and it’s the only thing keeping your product from blending into the median.
Your design system isn't agent-readable. Astryx shows what is
Meta open-sourced Astryx on September 13 — 150+ React 19 and StyleX components where CLI docs and MCP are the real interface. For solo founders, it's proof that agent-readable tokens beat longer prompts.
The prompt tax is why your AI product feels generic
Every extra line you prompt is a tax you pay in drift, review time, and a product that looks like everyone else. Here is how solo founders replace prompts with taste files agents can actually read.
Taste survives when you say 'you decide' — if your system lets it
One solo builder told Claude 'you decide' and shipped a design users thought was hired out. Here's what the three-day test says about taste, feel, and the system that holds it.
56% of tokens, 14% of spend: the split that should decide your AI stack
Vercel’s September gateway data shows open-weight models run 56% of tokens for 14% of spend while Anthropic still takes 64 cents per dollar. Here’s how solo founders use that split.
هندسة التصميم بالذكاء الاصطناعي: كيف تُسرّع فرق المنتجات دون خسارة الجودة
دليل عملي للنماذج اللغوية وأنظمة التصميم الآمنة للوكلاء وأتمتة سير العمل.
لماذا تركت وظيفتي لبدء شركتي الخاصة
لماذا تركت وظيفة مريحة لأعمل لحسابي — وما ربحته وما خسرته.
ما تعلمته من سنتي الأولى في العمل المستقل
أساطير الكفاح، دورات المال، الوحدة، زحف النطاق، والقاعدة الذهبية.
كيفية تصدير البيانات الوصفية من MDX لتحسين محركات البحث في Next.js
صدّر البيانات الوصفية من MDX لعناوين وأوصاف وروابط أساسية في Next.js.
Start Today.
Deciding to write every day. Not because I have something to say. Because I want to get better at saying it.
mosh desk
One person. Multiple tools. This is the machine.
When I became ramen profitable
Ramen profitable means covering the basics on your own terms — slowly, one step at a time.
The Founder Floor
The solo founder's biggest unlock in 2026 isn't a new tool. It's learning to operate on briefs and approvals instead of keystrokes.
AI-generated code is cheap. Production engineering is not.
Vibe coding gets you a prototype in hours. Making it reliable, secure, and maintainable takes a different kind of work — and AI has not made that work any faster.
Taste tools are the most important AI category in 2026
Half a dozen companies shipped taste infrastructure in a single window. That density tells you something about where AI building is headed.
Generative UI is a design problem
Interfaces that assemble themselves at runtime change what design means. Here is how design engineers should think about generative UI.
One-person startups are rewriting the rules of product taste
More than one in three new startups are solo founders. That changes what good design means — because taste can't be a team sport anymore.
عندما أصبحت مربحاً بما يكفي للرامين
الربح الكافي للرامين يعني تغطية الأساسيات بشروطك — ببطء وخطوة واحدة.
Connect
Feel free to contact me at hi@mosh.fyi