GlowInspo: When AI Dresses My Monday Anxiety

Source: Dev.to

The Problem

Remote workers, founders, creators, and consultants often experience rapid emotional shifts before 9 AM.

Example – 7:03 AM:

• Rain outside
• Back‑to‑back meetings
• Feeling overwhelmed

Instead of grabbing the first clean sweater, they need something that stabilizes them.

GlowInspo targets the decision‑fatigued morning crowd – women who want their clothing to support their mental state, not compete with it.

• Not about fleeting trends
• About operational energy
• A 2‑minute reset to move quickly between emotional states

What I Built

GlowInspo: 2‑minute AI‑powered emotional‑dressing reset

1. Check‑in – e.g., “overwhelmed Monday morning”
2. Energy selection – choose a goal: Stabilize, Soften, Brighten, Amplify
3. Claude Ritual – Arrival → Reframe → Direction
4. Gallery – 900+ Pinterest‑sourced outfits matched to the selected energy and tone
5. Memory – OOTD uploads create personal identity snapshots for future reference

Demo

• Live Demo: (link placeholder)
• Video walkthrough: (link placeholder)

Code

# Repository URL (replace with actual link)
git clone https://github.com/yourusername/glowinspo.git
cd glowinspo

# Install dependencies
pip install -r requirements.txt

# Run the Streamlit app
streamlit run app.py

A 2‑Minute Styling Ritual for Women Who Dress With Intention

GlowInspo is an emotion‑aware styling ritual designed for women juggling busy workdays, creative pursuits, and shifting internal moods.

• Purpose: Help you dress for the energy you want to bring into your day—not for fleeting trends.
• Benefit: Reduce decision fatigue while aligning your wardrobe with your emotional goals.

This version was created for the DEV Challenge to support women seeking more intentional, emotionally‑aligned dressing.

The Community

GlowInspo is designed for:

• Women in fast‑paced professional roles
• Creatives navigating identity and visibility
• Anyone who uses clothing as self‑expression

Our focus

Women who want to feel intentional—not overwhelmed.

Getting dressed shouldn’t feel like another task; it can be a reset.

How It Works

1. Check‑in: How are you arriving today?

2. Choose how you want to show up:

   • Soften
   • Stabilize
   • Brighten
   • Amplify

GlowInspo – Emotionally Aligned Outfit Guidance

GlowInspo is a lightweight, stateful Streamlit app that delivers curated outfit suggestions aligned with the user’s current emotional state. If you’re inspired, you can upload your own look to enrich the experience.

How I Built It

GlowInspo is built as a single‑page Streamlit interface that leverages session state to manage:

• Current recommendation set
• Feedback state (Inspired / Not Quite / Self‑Styled)
• Emotional memory history
• Controlled reshuffle logic

Session state prevents unnecessary recomputation and stabilises UI behaviour across interactions.

Architecture

Data Model & Tagging

Outfit data lives in a CSV file with the following columns:

Tagging Hierarchy

1. Primary filter → energy_bucket
2. Secondary ranking → tone‑overlap score

Tone overlap is computed on the fly:

filtered["match_score"] = filtered["tone_list"].apply(
    lambda tl: len(set(tl).intersection(set(tones)))
)

AI‑Assisted Tagging

Claude was used in a constrained, prompt‑based workflow to:

• Normalise the tone vocabulary across ~990 items
• Ensure each item receives 2–3 consistent tone tags
• Assign every item to one of the four defined energy buckets

The model was locked to a fixed tone taxonomy to avoid label drift, resulting in scalable yet emotionally consistent tagging.

Emotional Guidance Layer

Claude generates a structured output for each recommendation:

• Arrival
• Reframe
• Direction
• Energy
• Tone tags

The prompt enforces this format and restricts the allowed tone vocabulary, preserving system consistency.

Controlled Recommendation Logic

• Category‑aware diversification ensures varied suggestions.
• Recommendations stay frozen until the user explicitly requests a reshuffle.
• A manual reset is only possible via the “Not Quite” feedback, preventing infinite scrolling and reducing cognitive overload.

Identity Memory

User‑uploaded OOTDs (Out‑Of‑The‑Door looks) are stored in session state with:

• Mood
• Energy
• Composite identity label (e.g., “Tired + Amplify”)

When retrieving memory, the system energy‑matches prior successful identity states, which curbs unnecessary reshuffling and scroll fatigue.

Scaling & Future Extensions – Designing Human‑Centered AI Products

Persistent Storage

• Move emotional memory and uploaded images to cloud storage (S3, GCS, etc.).
• Store mood + energy snapshots in a database (PostgreSQL, Firestore).
• Enable cross‑session identity memory rather than session‑only state.

Richer Recommendation Engine

• Replace CSV filtering with a vector‑based similarity layer.
• Embed tone tags and outfit descriptions for semantic retrieval.
• Personalise suggestions based on historical energy patterns.

Generative Outfit Composition

• Leverage multimodal GenAI models to:
  • Generate full‑outfit combinations instead of single items.
  • Create personalised styling boards.
  • Suggest complementary pieces dynamically.

Future iterations could include image‑to‑image refinement, where users upload a single piece and receive AI‑generated styled variations.

Community Layer

• Opt‑in sharing of energy‑based looks.
• Surface “Popular Amplify Looks” or community‑curated energy boards.

---

GlowInspo is intentionally lightweight now, but its architecture is ready to grow into a fully‑featured, human‑centered AI styling platform.