Inspirati## Inspiration
Choosing what to wear every day often involves checking weather forecasts, calendars, and rummaging through closets. We wanted to eliminate this daily friction by creating an ambient, voice-driven personal stylist. VoguePulse AI bridges natural conversational intelligence with smart personal styling, letting users consult their wardrobe hands-free right inside the living room.
What it does
VoguePulse AI is an ambient wardrobe and styling assistant built as a self-hosted Model Context Protocol (MCP) server for Alexa+, backed by Amazon Bedrock.
- Wardrobe Inventory: Tracks cataloged items (tops, outerwear, trousers, shoes) and their style tags.
- Contextual Outfit Recommendations: Alexa+ calls the MCP server with the user's intended occasion (e.g., "formal interview", "casual weekend dinner") and current weather conditions.
- Generative AI Styling: Amazon Bedrock evaluates combinations from the available inventory and generates coordinated outfit suggestions with concise rationales.
How we built it
- Alexa+ Integration: Implemented a self-hosted MCP server adhering to the 2025-11-25 Streamable HTTP specification to expose tools directly to Alexa+ agent workflows.
- Cloud & AI (AWS Builder): Integrated Amazon Bedrock using
boto3to run inference models that evaluate color palettes, dress codes, and weather suitability. - Backend Service: Developed a lightweight, asynchronous Python service using FastAPI and Pydantic for schema validation and tool routing.
Challenges we ran into
Aligning the JSON tool-calling schemas with the exact Model Context Protocol (MCP) Streamable HTTP requirements and managing clean fallback logic for asynchronous LLM inference responses.
What we learned
We gained hands-on experience designing modular MCP servers, writing structured tools for agentic systems, and orchestrating generative inference via Amazon Bedrock.on
What it does
How we built it
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for amazon hackthon
Built With
- alexa+
- amazon-bedrock
- aws-sdk-(boto3)
- fastapi
- model-context-protocol-(mcp)
- python
- rest-api
