Raluca Florescu

A Proposal for AI-Assisted Retail Applications

Modern shopping applications are often designed around quantity rather than clarity. Instead of helping users discover products they genuinely enjoy, many platforms overwhelm them with endless scrolling, repeated items, poor filtering systems, and disconnected recommendations.

The result is frustration, decision fatigue, and users feeling disconnected from the products they are seeing.

This issue is especially visible in clothing applications, where people are not only purchasing products, but trying to build outfits, aesthetics, comfort, confidence, and visual coherence.

The Problem With Current Shopping Systems

Most shopping platforms repeatedly show users products they already ignored, disliked, or would never realistically purchase.

Users often:

• scroll through hundreds of irrelevant items
• repeatedly see products they dislike
• lose focus because unwanted products dominate attention
• feel visually overwhelmed
• struggle to imagine combinations
• become tired before finding something they genuinely enjoy

Instead of learning from rejection, many applications simply continue pushing inventory endlessly.

This creates an exhausting shopping experience.

Preference-Based Visual Selection

Shopping applications should allow users to visually reject products they do not want to see again.

For example:

• swipe left or press X for rejection
• permanently reduce similar recommendations
• remove unwanted silhouettes, cuts, or styles
• refine recommendations dynamically over time

This creates a calmer and more intentional experience.

The user should not need to repeatedly filter out the same unwanted products every session.

Building a Personal Preference Memory

Applications should remember:

• preferred sizes
• preferred colors
• favorite textures and fabrics
• preferred silhouettes
• disliked cuts
• transparency tolerance
• comfort preferences
• desired clothing proportions
• oversized vs fitted preferences

If a user consistently rejects oversized shirts, extremely short tops, or transparent fabrics, the application should naturally adapt instead of continuing to display similar products.

The goal is not simply recommendation repetition, but intelligent preference refinement.

Shopping as Peaceful Exploration

Users should have the ability to explore products without pressure.

Current shopping applications often interrupt users constantly with:

• checkout popups
• repetitive purchase prompts
• aggressive notifications
• forced urgency

However, many users browse in discovery mode, not immediate buying mode.

The application should allow people to:

• collect ideas calmly
• compare products visually
• build temporary selections
• revisit preferred items later
• think without interruption

The shopping process should feel relaxing rather than cognitively exhausting.

Reducing Visual Noise

One of the largest problems in online shopping is excessive visual clutter.

Showing users hundreds of products at once often reduces emotional connection and increases decision fatigue.

Instead of presenting massive catalogs continuously, applications could:

• show smaller curated selections
• allow quick rejection of unwanted items
• prioritize preference learning
• refine recommendations dynamically

This creates a stronger focus on products the user actually enjoys seeing.

“What If?” Product Curation

Shopping should not feel like sorting through digital garbage.

After removing products users dislike, the application should preserve products that create curiosity, inspiration, or aesthetic interest.

The goal becomes:
“What if this works?”
“What if this fits my style?”
“What if this completes an outfit?”

This emotional curiosity is more valuable than endless inventory exposure.

AI-Assisted Outfit Composition

Users often struggle not because they dislike products individually, but because they cannot imagine how products work together.

Applications should actively help users create outfits by suggesting:

• matching pants or trousers
• compatible tops
• shoes
• jewelry
• bags
• coats
• makeup or hairstyle inspiration
• complete visual combinations

This transforms shopping from isolated purchasing into guided aesthetic construction.

Comfort, Elegance, and Real Wearability

Many users are searching for clothing that feels:

• elegant
• soft
• comfortable
• feminine
• structured without restriction
• visually coherent

Current fashion systems often oscillate between:

• oversized shapeless clothing
• excessively tight clothing

Many users instead prefer balanced silhouettes:

• soft structure
• movement
• body-aware cuts
• comfortable fabrics
• material presence without stiffness

Applications should help users discover products based on how clothing behaves in real life, not only through static product images.

Discovery Mode

While preference learning is important, users should still have the possibility to explore unexpected styles.

Applications could include a separate “Discovery” mode where users voluntarily explore:

• experimental aesthetics
• unusual products
• trend-based pieces
• visually different combinations

However, discovery should remain controlled and interactive rather than overwhelming.

Conclusion

The future of shopping applications may depend less on displaying more products and more on reducing cognitive overload.

Instead of functioning as endless digital catalogs, shopping platforms could become intelligent systems that:

• learn user preferences
• reduce unwanted repetition
• simplify visual selection
• guide outfit construction
• preserve curiosity without overwhelming the user

The goal is not only increasing sales.

The goal is helping people feel calm, understood, visually confident, and emotionally connected to what they choose to wear.