Cartivo

Your AI shopping copilot. Analyze trends, videos, and style data to generate confidence-scored, compatible bundles for smarter, interactive fashion decisions

Comment

Inspiration

Shopping online can be overwhelming: from conflicting trends to uncertainty about fit, style, or compatibility. The avalanche of e-commerce sites has further compounded this problem. We wanted to create an AI assistant that gives users confidence-backed, curated fashion and lifestyle recommendations, transforming data, trends, and visuals into actionable choices.

What it does

Cartivo analyses videos, trends, and research in real-time to generate confidence-scored, compatible bundles. Users can explore outfits, verify trends, track style evolution across scenes, collaborate with teams on a visual board, and even export PDF reports of curated bundles. It’s interactive, reliable, and personalised. No more visiting multiple e-commerce sites to source the best deals and compare options. With Carvito, you have an autonomous agent that does this for you.

How we built it

We combined:

  • Cartivo AI engine for compatibility scoring and bundle generation
  • You.com API for live trend verification and citation-backed insights
  • Perfect Corp API for visual fit and style validation
  • Miro API for collaborative, real-time bundle planning
  • DeepGram Autonomous voice execution agent
  • Postgres UI for UI components (Cards, Charts, MultiSelect, etc.)
  • Foxit PDF API for exporting sharable reports
  • Sanity CMS to manage structured content and product data
  • Figma Make UI and wireframes
  • Replit AI to build the web app
  • KiloCode to accelerate development

We orchestrated these services to work in real time, merging AI reasoning with structured content and multi-user collaboration.

Challenges we ran into

  • Integrating multiple APIs while maintaining real-time updates and low latency
  • Building a narrative-aware video analysis pipeline for style evolution
  • Mapping complex AI insights into simple, intuitive visual widgets for Miro
  • Ensuring confidence scores were meaningful and easy to interpret

Accomplishments that we're proud of

  • Fully interactive, multi-user Miro integration for collaborative outfit planning
  • Real-time, evidence-backed recommendations using You.com insights
  • StoryLens video analysis to track style evolution and key moments
  • Confidence Index™ system that merges AI compatibility, visual fit, and trend data
  • A hackathon-ready demo that’s visually compelling, technically robust, and fully functional

What we learned

  • Combining AI reasoning with real-time web data and structured content creates trustworthy and actionable recommendations
  • Multi-user collaboration requires careful handling of events, latency, and updates
  • Users value context-aware insights (e.g., narrative-driven fashion trends) more than static suggestions
  • Integrating multiple APIs can produce powerful, cross-domain solutions if orchestrated carefully

What's next for Cartivo

  • Expand support for more video platforms and influencer content to improve trend detection
  • Add motion-based wearability scoring for dynamic outfit recommendations
  • Enhance Miro integration with gamified bundle workshops
  • Launch a consumer-facing app for individual shoppers while keeping team collaboration features
  • Explore AI-generated styling suggestions for lifestyle products beyond fashion

Built With

  • cocreate
  • deepgram
  • figma
  • foxit
  • kilocode
  • miro
  • opencorp
  • postgresui
  • react
  • replit
  • sanity
  • you.com
Share this project:

Updates