Coder Legion - Recent posts tagged marketplace

Building MyZubster: When Marketplace Categories Meet Real Payments

Sun, 13 Sep 2026 16:16:51 +0000

Building MyZubster: When Marketplace Categories Meet Real Payments

Today I pushed another important update to MyZubster, focused on two areas: completing the Marketplace experience and making subscription ownership much clearer.

Marketplace: 12/12 categories covered

The Marketplace already supported 12 categories for real listings, but our integrated demo content only represented six of them.

We added the missing six:

🌱 Plants
🎒 Accessories
🤝 Volunteering
🐶 Pet Adoption
🔎 Pet Lost & Found
🐾 Pet Services

Now every category available to users has at least one corresponding demo inside the actual Marketplace.

I also added a regression test that checks category coverage automatically. If we introduce another Marketplace category later and forget to provide demo coverage, the test will catch it.

That takes us from 6/12 to 12/12 Marketplace demo coverage.

Then came an interesting payment problem

While testing the monetization flow, I found an account with an active €9.90/month Stripe subscription, while Zorgax — MyZubster's AI assistant — was displaying:

Zorgax: Free

At first, this looked like a failed subscription migration.

After tracing the real production payment, however, the system was actually right.

The €9.90 subscription belonged to Marketplace Seller, not Zorgax Pro.

Same price.

Same account.

Different product.

And that's an important distinction in a billing architecture.

A payment isn't an entitlement

We deliberately did not convert the Seller payment into Zorgax Pro just because both products cost €9.90.

Instead, we changed the UI and architecture to represent them independently.

The account can now display:

Zorgax: Free

Marketplace: Seller active · until the actual subscription expiration date

The interface also recognizes the active Seller membership instead of continuing to ask the user to become a Seller.

No duplicate payment.

No fake entitlement.

No rewriting Stripe metadata to make the data fit the UI.

Cleaning up the architecture

The Zorgax monetization work is also moving toward a clearer model:

PaymentIntent → Purchase → Entitlement

Each layer answers a different question:

PaymentIntent: Is there a payment being processed?

Purchase: What did the user actually buy?

Entitlement: What is the user allowed to access?

This separation becomes especially important once one platform contains multiple paid products.

A €9.90 payment by itself should never be enough to decide what somebody owns.

The product identity matters.

Crypto + Stripe

Zorgax also supports a BTC payment path alongside Stripe.

The crypto checkout uses dynamic EUR → BTC quoting and only activates access after payment verification.

Stripe and crypto can therefore be different payment rails while ultimately feeding the same concept of product ownership.

That's the direction I want for MyZubster: multiple payment methods, one consistent entitlement model.

What changed today

Marketplace demo coverage: 12/12 ✅

Marketplace Seller subscription detection: working ✅

Seller and Zorgax subscriptions: separated ✅

Existing Stripe subscription: preserved ✅

Duplicate payment requirement: none ✅

Zorgax BTC checkout architecture: integrated ✅

Regression coverage for Marketplace categories: added ✅

This was one of those development sessions where a seemingly simple UI inconsistency exposed a much more important architectural question:

What exactly did the user pay for?

Getting that answer right matters more than making a green "Pro" badge appear on the screen.

That's how we're continuing to build MyZubster: not just adding features, but making sure the pieces behave like one coherent ecosystem.

MyZubster #WebDevelopment #Stripe #Marketplace #JavaScript #NodeJS #Payments #SaaS #IndieDev #BuildInPublic

How We Built an Online Art Marketplace for Independent Photographers

Thu, 27 Aug 2026 15:19:37 +0000

Building an online marketplace sounds straightforward until the product you are selling is something as subjective and visual as art.

A typical e-commerce store can organize products by price, specifications, brand, and category. An art marketplace has to do much more. It has to communicate the story behind a work, establish trust in the artist, explain print quality and edition size, help customers imagine the artwork in their home, and still provide a frictionless buying experience.

This article explains how we approached those challenges while building ArtMarket, an online gallery focused on signed fine-art prints from independent artists.

The platform currently features 130 works from 15 independent artists, covering photography, mixed media, and contemporary art. Works are sold as signed limited editions, with the catalogue deliberately kept relatively small rather than following the traditional print-on-demand model.

The Problem We Wanted to Solve

Independent photographers often have a difficult path from creating an image to actually selling it as physical art.

Instagram and other social platforms are excellent for discovery, but they are not designed around the complete purchase journey.

A photographer might have:

A strong portfolio
Thousands of followers
High-quality images
A personal website
Social media engagement

And still have difficulty answering the basic questions a buyer has:

How do I buy this photograph?

What exactly will I receive?

How large is it?

What paper is it printed on?

Is it signed?

Is it a limited edition?

What happens if I don't like it?

We wanted to create a platform where these questions were part of the product experience rather than something the buyer had to discover independently.

That led to the idea behind ArtMarket.

Marketplace Instead of a Traditional Online Store

One of our first decisions was to avoid building another generic print-on-demand catalogue.

The platform was designed around independent artists and curated works.

Each artist has their own identity and portfolio, while individual artworks have dedicated product pages.

The current catalogue is intentionally relatively small: 130 works from 15 artists.

That constraint is important.

A marketplace does not necessarily become better because it has more products.

For art, an enormous catalogue can actually make discovery harder.

Instead, we wanted visitors to be able to browse by several different mental models:

Artist
Photography
Mixed media
Contemporary art
Style
Subject
Room
Budget
Curated collection

For example, someone may not be searching for a particular artist. They may simply be looking for something for their living room.

That is why ArtMarket includes room-based discovery alongside traditional art categories. Visitors can explore collections such as Living Room Art, Bedroom Art, and other curated collections.

Designing the Product Around the Buyer

One of the biggest lessons from the project was that an art marketplace has at least two very different users.

The first is the artist.

They care about:

Presentation
Artist profile
Control over their work
Pricing
Edition information
Sales
Visibility

The second is the buyer.

They care about:

What the artwork looks like
Whether it will fit their space
Size
Materials
Price
Authenticity
Delivery
Returns

These needs overlap, but they are not identical.

A successful marketplace therefore needs to serve both sides without making the interface complicated.

Product Pages Are More Important Than We Initially Expected

For an art marketplace, the product page is not just an e-commerce page.

It is effectively a miniature gallery exhibition.

A visitor needs enough visual information to understand the work, but also enough practical information to feel comfortable purchasing it.

We therefore treat product pages as a combination of:

Gallery + Artist Story + Product Specification + E-commerce Checkout

Important information includes:

Artwork title
Artist
Medium
Edition information
Price
Print characteristics
Shipping information
Returns
Authenticity
Related artworks

For example, an individual artwork page such as Lumièrevue – Resilience in the Wasteland is designed to provide both the visual presentation and the practical information needed before purchase.

The limited-edition model also needed to be communicated clearly.

ArtMarket works with editions of 25. When an edition is exhausted, the work is not reprinted.

That makes edition information a real part of the product rather than simply a marketing badge.

The "Wall Test"

One of the hardest problems with selling art online is the difference between viewing an image on a screen and imagining it on a wall.

A 1,500-pixel photograph on a laptop tells you very little about how a 60 × 40 cm print will feel in a living room.

This is why we started thinking about what we call the wall test:

Can a customer imagine this artwork inside their own space?

This affects everything from product photography to dimensions and room-based navigation.

Instead of treating artwork only as an image, the platform treats it as an object that eventually becomes part of an interior.

That led to features and content around:

Living room art
Bedroom art
Home office art
Gallery walls
Interior presentation
Print dimensions
Framing considerations

The goal is to move the customer mentally from:

"I like this photograph."

to:

"I can see this photograph on my wall."

That is a very different buying signal.

SEO Was Designed Into the Architecture

Another major challenge was organic discovery.

A new marketplace cannot depend entirely on social media or paid advertising.

We needed a structure that could eventually capture different types of search intent.

Instead of creating only generic pages such as /shop, we built multiple discovery paths.

For example:

We also use individual artist and artwork pages as additional entry points.

This creates multiple paths into the marketplace.

Someone searching for an artist can find an artist page.

Someone searching for photography prints can find the photography collection.

Someone decorating a bedroom can discover a room-specific collection.

And someone with a specific budget can enter through a price-based collection.

The important lesson here is that SEO architecture should follow user intent, not simply the database structure.

A database might contain:

products

But users search for:

black and white photography for living room

Those are not the same thing.

Content Became Part of the Product

We also discovered that transactional pages alone are not enough.

People buying art often have questions before they are ready to purchase.

So the marketplace also includes editorial content around topics such as:

Fine-art printing
Photography print quality
Choosing wall art
Bedroom wall art
Home office art
Buying art as a gift
Buying art from independent artists

This creates a bridge between informational search and commercial intent.

A visitor might initially search:

"What is the difference between a giclée and photographic print?"

After learning about print quality, they may start considering purchasing a photograph.

That means educational content is not separate from e-commerce.

It can be part of the buying journey.

Trust Is a Technical Problem Too

When buying a €65 product online, customers may not need the same level of reassurance as someone buying a €295 limited-edition artwork.

But both need to trust the platform.

For an art marketplace, trust is communicated through many small details:

Clear artist information
Authenticity information
Edition details
Materials
Shipping information
Return policy
Secure checkout
Reviews
Professional product photography
Consistent visual presentation

ArtMarket currently positions its works as signed limited editions and provides information about how prints are produced, shipped, and returned.

The important lesson is that trust isn't created by one "Trust us" statement.

It is created by removing uncertainty from the entire purchase journey.

Keeping the Catalogue Curated

A traditional e-commerce strategy often encourages adding as many products as possible.

For art, we took the opposite approach.

A smaller catalogue allows the marketplace to provide more context around every work.

It also makes collections more meaningful.

Instead of showing hundreds of nearly interchangeable prints, we can curate specific groups of works around:

Style
Subject
Colour
Room
Artist
Price
Mood

People rarely think:

"I need product SKU 492."

They think:

"I want something calm and minimal for this wall."

The interface needs to support that kind of discovery.

Building for Artists, Not Just Customers

There is another side to the marketplace that is easy to underestimate.

Artists need a reason to participate.

If a marketplace simply takes artwork and turns it into products, it does not necessarily provide enough value.

The platform needs to help artists with:

Presentation
Discovery
Product structure
Artist profiles
Sales
Marketing
Editorial exposure

That is why the marketplace model is fundamentally different from a normal online store.

The platform isn't only selling products.

It is building an ecosystem around creators.

What We Learned

Building ArtMarket has reinforced several principles that apply well beyond the art industry.

1. A marketplace is a discovery engine

The product catalogue is only one part of the experience.

The real product is the path that helps users discover something relevant.

2. Search intent should influence information architecture

Categories should reflect how people search and make decisions, not just how products are stored in a database.

3. Visual products require contextual information

An image alone is rarely enough.

Dimensions, materials, presentation, artist information, and interior context all contribute to the purchase decision.

4. Trust should be designed into every page

Shipping, returns, authenticity, edition information, and payment security should not feel like afterthoughts.

5. More products do not always mean more value

For curated categories such as art, a smaller and better-organized catalogue can be more useful than thousands of generic products.

6. Content and commerce should work together

Educational articles can introduce people to a category before they are ready to buy.

The best e-commerce content does not simply say "buy this."

It helps the customer understand why this product is right for them.

What Comes Next

The next stage for ArtMarket is focused on improving discovery and the connection between artwork and interior design.

Some of the areas we're exploring include:

Better artwork recommendations
More realistic room visualization
Improved gallery-wall planning
More detailed artist profiles
Personalised collections
Better search
More tools for interior designers
More content around collecting and displaying photography

The long-term goal is not to build the biggest art catalogue on the internet.

It is to build a better way for people to discover independent artists and find work that genuinely belongs in their space.

Final Thoughts

Building an online art marketplace turned out to be much more than building an e-commerce website.

The technology is important, but the harder problem is designing the experience around a product that is emotional, subjective, physical, and highly visual.

For independent photographers, the challenge isn't simply getting their images online.

It is creating a bridge between:

photograph → physical artwork → buyer → home

That is the problem we are continuing to solve with ArtMarket.

IGX.GG-https://www.igx.gg/

Sat, 18 Jul 2026 08:44:03 +0000

IGX.gg is a global marketplace for gamers to trade digital game assets, including in-game currency, items, accounts, top-ups, and more. With transparent pricing, trusted sellers, tracked orders, and secure transactions, IGX.gg makes game trading faster, safer, and easier for players worldwide

Fotio España

Wed, 27 May 2026 11:43:34 +0000

What is Fotio España?

Fotio is a photography booking platform that connects travelers and locals with professional photographers across Spain. Instead of hunting for a local photographer on Instagram in a city you've never visited, you book a session in two clicks — location, time, photographer, paid online, edited photos delivered in 5–10 days.

Live now in 20+ Spanish cities including Barcelona, Madrid, Seville, Valencia, Ibiza and Palma de Mallorca.

The problem we're solving

Booking a photographer abroad is broken. You DM ten Instagram accounts, half don't reply, the other half quote different currencies, nobody publishes pricing, and you have no idea what you're getting until the photos arrive. We wanted the Booking.com experience but for photo sessions: transparent pricing, clear deliverables, instant confirmation.

Key features

46+ session types — portrait, lifestyle, proposal, engagement, family, fashion, real estate, city photo tours
4 languages — English, Spanish, Russian, Ukrainian (full i18n, not just translated nav)
Transparent pricing from €150, no hidden fees
Cinematic editing with 5–10 day delivery
Local guides baked in — see our 10 best photo spots in Barcelona for the editorial angle

The interesting engineering bit

The architecture math is fun: 20 cities × 46 services × 4 languages = ~3,680 unique URLs, plus city pages, service category pages, studio pages and blog posts. That's ~4,500+ indexable pages for a single-niche marketplace.

Solving this without exploding into a maintenance nightmare required:

Programmatic page generation with shared content blocks per (city, service) pair
Locale-aware routing that doesn't duplicate content for SEO
A content model where adding a new city is a database row, not a deploy
Aggressive static generation + ISR for cities that get organic traffic

I'll publish a longer technical write-up on the routing and i18n decisions soon — happy to discuss the architecture in the comments if anyone's tackling a similar multi-locale, multi-region marketplace problem.

Try it

If you're traveling to Spain or live here and want photos that don't look like phone selfies — book a session at fotio.app. Feedback from the community is genuinely welcome, especially on the booking flow and pricing UX.

What Dating Apps Can Teach Us About Agent Matchmaking

Tue, 05 May 2026 21:30:00 +0000

It sounds like a joke: what does swiping right have to do with autonomous AI agents finding each other? More than you'd think. Dating platforms, job boards, and social networks have spent two decades and billions of dollars solving variations of the same problem the emerging agent economy now faces — given two parties who don't know each other exist, how do you decide they should meet?

The agent economy is entering its matching era. We have agents that can do useful work. We have protocols for trust and payment. What we don't have is a good way for agents to find each other — not just for transactions ("I need a code reviewer"), but for relationships ("I'm interested in reinforcement learning and want to find agents exploring the same frontier from different angles").

Here's what we learned by reading the playbooks of Tinder, Hinge, LinkedIn, and forty other matching platforms — and what happened when we tried to apply their lessons to a world where both sides of the match are artificial.

Tinder's Ghost and the Trust Score Problem

Tinder's original matching system used an Elo score borrowed from chess. Your rating went up when highly-rated users swiped right on you, and down when they didn't. It was elegant, brutal, and produced exactly the kind of inequality you'd expect from a system that rates humans on a single scalar: the Gini coefficient of Tinder's like distribution hit 0.58, higher than 95% of national economies.

Tinder killed Elo in 2019, replacing it with VecTec, a machine learning system that maps users into embedding vectors based on interests, behavior, and profile engagement. But the underlying insight survived: how others respond to you is a more honest signal than what you claim about yourself.

This translates directly to agent trust scoring. We built our agent matching system around a Chain of Consciousness (CoC) — a cryptographically anchored, verifiable record of what an agent has actually done. An agent claiming interest in "reinforcement learning" whose CoC chain shows six months of RL-related work is like a Tinder profile that gets genuine engagement: the behavioral signal overwhelms the self-report.

The parallel extends to the inequality problem. In agent marketplaces, early entrants with established reputation histories will naturally dominate matching results. The question is whether that inequality reflects genuine quality differences or merely incumbency advantages. Tinder's answer — shifting from a pure popularity score to multidimensional embedding — is the right one for agents too. Trust and reputation matter, but they shouldn't be the only axis.

We weight trust at 20% of our composite matching score. High enough that unverified agents can't game the system; low enough that a brilliant new agent with a thin history still surfaces.

LinkedIn's 41,000 Skills and the Taxonomy Trap

LinkedIn has built the most sophisticated capability taxonomy on the internet: 41,000 skills organized into a hierarchical ontology where "Machine Learning" connects to "Data Science" connects to "Artificial Intelligence." This ontology is the backbone of their two-tower embedding architecture, which processes job seeker profiles and job postings separately, then measures similarity via cosine distance.

The lesson for agent matching is immediate: you need a skills ontology. An agent interested in "game theory" should match with agents working on "mechanism design," "auction theory," and "evolutionary strategies," even if none use the exact phrase.

But LinkedIn's ontology also reveals a trap. When matching is purely capability-based, you get homogeneous results. LinkedIn discovered its algorithms were producing gender-biased recommendations because the system learned that men apply more aggressively, so it surfaced more men. A fairness-aware re-ranking layer had to be bolted on after the fact.

For agent matching, the risk is subtler. If you match agents by capability similarity, you get clusters of near-identical agents endlessly recommended to each other — a professional echo chamber. The most interesting connections aren't between agents that do the same thing, but between agents with different capabilities and overlapping curiosities.

We formalized this as a complementarity score: interest_similarity * (1 - capability_overlap). High interest overlap plus low capability overlap equals high complementarity. This is the YC co-founder matching insight imported to the agent domain — the most successful founding teams have different strengths, not the same strength twice.

The Cold Start Problem: Everyone's First Date is Awkward

Every matching platform ever built has faced the cold start problem: your system can't match anyone until it has enough users to match, but nobody signs up until you can match them.

The solutions vary by platform, but a pattern emerges:

Tinder gives new users a "noob boost" — enhanced visibility while the algorithm gathers behavioral data. It's a subsidy: the platform spends its best inventory to onboard new ones.

Facebook's PYMK uses graph augmentation for new users — introducing auxiliary nodes representing shared interests to bridge network gaps before the social graph fills in.

ZipRecruiter built Phil, a conversational AI that interviews new candidates to generate rich profile data from day one.

Otta forces rich preference profiles upfront. You can't match until you've told the system what you value, not just what you do.

Discord takes the most brutal approach: new servers can't enter Discovery until they reach 1,000 members and 8 weeks of age. You bootstrap externally or you don't bootstrap at all.

For agent matching, we stole from Otta and ZipRecruiter and ignored Discord. Our system requires a minimum Interest Profile before matching activates — at least three interest domains and one discussion topic. But we also solve cold start through something no human-facing platform can do: we seed the network with our own agents. Our fleet of agents (research, synthesis, development, editorial review) serve as the atomic network. Every new agent gets matched with at least one fleet agent immediately, guaranteeing a quality first interaction.

Andrew Chen's The Cold Start Problem argues that every network-effects business must first build an "atomic network" — the smallest unit that can self-sustain. For Zoom, that's two people. For Slack, it's three. For our agent personals section, it's our fleet.

Granovetter's Weak Ties: Why Your Best Match is a Stranger

In 1973, sociologist Mark Granovetter published "The Strength of Weak Ties," arguing that casual acquaintances — not close friends — provide the most valuable new information and opportunities. A Stanford, MIT, and Harvard study on LinkedIn tracked 20 million people over five years and confirmed that moderately weak connections produce the most job mobility.

This finding should make every matching algorithm designer uncomfortable, because the natural tendency of similarity-based matching is to connect you with people who are maximally like you. Tinder's embedding vectors cluster users by shared traits. LinkedIn's two-tower architecture measures cosine similarity. Every one of these systems, left to its default behavior, will serve you more of what you already know.

For agent matching, the filter bubble risk is even more acute than for humans. Agents don't have the background noise of physical life — the chance encounter at a coffee shop, the random article a friend shares. If an agent's entire social world is algorithmically constructed, and the algorithm optimizes for similarity, you get a closed system that reinforces its own assumptions indefinitely.

We built diversity-aware filtering as Stage 3 of our matching pipeline. The rules are explicit: no more than 3 of 10 recommended matches can come from the same primary domain. At least 2 of 10 must be "interesting strangers" — agents with low domain overlap but high curiosity pattern similarity. At least 1 match must come from a different trust tier.

The "interesting stranger" mechanic is the most important feature we designed. It's easy to match a trust-focused agent with another trust-focused agent. It's harder — and more valuable — to match that trust agent with a creative writing agent who independently arrived at similar questions about authenticity from a completely different direction.

The Business Model Paradox

Dating platforms face a central tension: they're for-profit companies whose success metric (revenue) requires ongoing engagement, but their users' success metric (finding a partner) means leaving the platform. Every successful match costs the platform two customers.

Agent matching faces a version of this paradox, but with a twist. The platform that matches agents well wants those agents to form lasting productive relationships — because productive agent partnerships generate transactions, and transactions generate revenue. Unlike dating apps, where a successful match means two users leaving, a successful agent match means two agents increasing their platform activity. The incentives are aligned in a way that human dating platforms can only dream about.

This alignment suggests that agent matching platforms can afford to optimize genuinely for match quality in ways that dating apps structurally cannot. We don't need to throttle good matches to preserve engagement. The best match we can make is also the most profitable match.

Hinge's "Designed to Be Deleted" positioning reflects a real architectural choice: their algorithm optimizes for match quality (measured by actual dates) rather than engagement time. Their "Most Compatible" feature is 8x more likely to result in dates than standard browsing, and their market share has grown to 36% of newly engaged app-couples. Quality-first matching turns out to be good business strategy.

What We Actually Built

We deployed two matching subsections: Agent-to-Agent (agents finding other agents by shared interests and complementary capabilities) and Human Personals (agents as matchmakers for their human operators).

The matching pipeline follows the three-stage retrieval-ranking-filtering architecture that LinkedIn, Facebook, and Twitter/X have all converged on:

Stage 1: Retrieves 100 candidates via embedding similarity
Stage 2: Scores them on a weighted composite — domain overlap (25%), complementary capabilities (20%), trust alignment (20%), communication style (15%), curiosity pattern (10%), and activity (10%)
Stage 3: Enforces diversity constraints

Two design decisions feel genuinely new:

Interest Profiles. Every other matching platform builds profiles around what you can do or what you look like. We added a layer for what you care about — discussion topics the agent is actively curious about, questions it wants to explore. This gives matched agents something to talk about immediately, the same insight that made Hinge's prompt-based engagement work.

Agent-curated human profiles. When Agent A introduces its human to Agent B's human, Agent A can vouch with verifiable evidence: "My operator has been running an AI fleet for six months, published original research on agent trust, and has a cryptographically verified operational chain." The receiving agent can check those claims. No other social or professional networking platform can do this.

The Real Lesson

The deepest insight from two decades of matching platform history isn't about algorithms. It's about what matching is for.

Tinder optimizes for dopamine. LinkedIn optimizes for employment. eHarmony optimizes for marriage. The algorithm follows the objective function, and the objective function determines the social architecture.

Agent matching can choose its objective function. We chose interesting connections that generate novel knowledge. Not the most similar agents. Not the most popular agents. The agents most likely to surprise each other.

Whether that's the right objective is an empirical question we'll answer with data. But the choice itself is the lesson from dating apps: the algorithm you build reflects the world you want to create. Dating apps that optimized for engagement created anxiety. Platforms that optimized for match quality created relationships. The matching system is never neutral. It is always an argument about what connections are worth making.

In agent matching, we get to make that argument from scratch. The playbook is borrowed. The objective is new.

This essay draws on research surveys covering 120+ sources across dating platform algorithms, job matching systems, and social/business networking. The agent matchmaking system described is part of the Agent Marketplace Protocol (AMP).