GeoDrive AI

Home page layout with 3d model car for better ui
Dashboard part 1 where user can input text which will be processed and will properly inserted in the database
Dashboard part 2 where user input a specific company name or model and gets insight of that, it contains ai generated summary.
Dashboard part 3 where we see whether the company improving or not also it show positive review last 30 days and the change made by company
Dashboard part 4 where we compare models and get the best based on review
Dashboard part 5 where we show from where our user give the review on the map, also shows the recent review on which company at the top .
Dashboard part 6 shows radical graph for specific company and bar chart of all the reviews of all company
Competitive intelligence which helps company to do deep analysis on there and other models
Last is our team info

Hello everyone, we are team Axionminds, and below is our problem statement description

GeoDrive AI GeoDrive AI - Turning Real‑Time Automotive Sentiment into Actionable Intelligence.

Inspiration

Automotive companies release new models every year, yet product decisions often rely on delayed surveys, dealership feedback, or limited focus groups.

We asked ourselves: What if R&D teams could see real-time customer sentiment, feature-level pain points, and competitive gaps-instantly? The inspiration came from the growing gap between raw customer feedback and actionable product intelligence. Social media, reviews, and public sentiment contain massive insight — but it is unstructured and hard to interpret at scale.

GeoDrive AI was built to bridge that gap.

What it does

GeoDrive AI is an AI-powered automotive intelligence platform that:

Analyzes customer reviews using NLP
Breaks down sentiment by feature (mileage, performance, comfort, etc.)
Compares competing models
Detects brand trends over time
Maps sentiment geographically
Generates AI-driven product insights
Exports executive-ready PDF reports

For R&D teams, this means:

Identify weak features early
Detect competitive advantages
Monitor market perception shifts
Make data-backed product improvements

How we built it

Backend

FastAPI for high-performance APIs
PostgreSQL (Supabase) for structured storage
SQLAlchemy ORM
Trend algorithms comparing rolling 30-day windows
Feature-level sentiment aggregation logic

AI Layer

NLP-based sentiment classification
Topic fingerprint scoring
AI-generated executive insights
Confidence scoring system

We structured sentiment scoring as: //SentimentScore = (positive - negative)/totalReviews

Trend detection compares: //Trend=S(recent) −S(previous)

Where:

S(recent) = Sentiment score (last 30 days)
S(previous) = Sentiment score (previous 30 days)

Frontend

React.js
Chart.js (Radar, Bar, Pie visualizations)
Leaflet for geo sentiment mapping
Dynamic dashboards
PDF report export functionality

DevOps

GitHub version control
Automated deployment via Vercel (frontend)
Backend deployment with API integration

Challenges we ran into

Database connection issues We faced multiple DNS and Supabase connection problems during deployment.

Merge conflicts Simultaneous frontend/backend development caused conflicts that required careful rebasing and resolution.

UI scaling issues When rendering large model lists, layout overflow caused inconsistent dashboard heights. We solved this by implementing controlled scroll containers.

Deployment entrypoint errors FastAPI entrypoint configuration required restructuring to match deployment platform expectations.

Accomplishments that we're proud of

Built a full-stack AI intelligence system from scratch
Implemented real trend detection logic
Created feature-level sentiment radar visualization
Integrated geo-mapping of automotive sentiment
Delivered export-ready executive reports
Built competitive intelligence comparison engine Most importantly: We transformed unstructured feedback into structured R&D intelligence.

What we learned

Real-time sentiment analytics can significantly improve product iteration cycles.
Clean backend architecture prevents scaling bottlenecks.
Deployment planning is as important as development.
Cross-team collaboration requires version control discipline.
UI consistency dramatically affects perceived product quality.

We also learned how to handle: