-
The first page that explains your vehicle history and similar model stats
-
Second page helps you find your next car using AI
Inspiration
Inspiration: One app to tell a vehicle's complete story — factory to driveway.
What it does
What it does: Streamlit app with vehicle narrative, warranty tracking, journey map, AI sentiment, and an AI vehicle recommender.
How we built it
How we built it: Snowflake + Cortex Code. Semantic view over 14 tables, tested with Cortex Analyst, built the app and AI features — all inside Snowsight.
Challenges we ran into
Challenges we ran into: Learning Snowflake. Fan-out joins. Case sensitivity. AI function return types. LLM model deprecation.
Accomplishments that we're proud of
Accomplishments that we're proud of: Full build without leaving Snowflake. AI-powered sentiment and recommendations on real data.
What we learned
What we learned: Semantic views are iterative. VQRs matter most. CoCo makes the build loop fast.
What's next for A Vehicle Story
What's next for A Vehicle Story: Predictive warranty modeling, defect alerts, dealer analytics.
Built With
- python
- semantic-view
- streamlit
