I have created simple INDIAN-INFRA-AI-INSIGHTS Streamlit app 
I’m thrilled to unveil Indian Infra AI Insights, a Streamlit & Snowflake-powered AI application that combines data, AI, and innovation to tackle India’s infrastructure challenges.
From predicting accident severity, preventing recommendations, forecasting petrol demand, and analyzing air quality, this platform is a one-stop solution for actionable insights.
Technologies Used/Learnt/Explored
WeatherAPI.com,Snowflake,streamlit,SNOWFLAKE.ML.FORECAST,TASKS,
SNOWFLAKE-RAG,SNOWFLAKE.CORTEXT,COMPLETE, pdf_text_chunker ,Dynamic tables,SNOWFLAKE.CORTEX.PARSE_DOCUMENT,Hosted on Streamlit cloud, LangChain, Google Gemini , Data sharing between SF accounts and many GA features
Top Features of the App
Accident Severity Prediction using AI models.
Real-time Air Quality Index (AQI) visualization for Indian cities.
Petrol Demand Forecasting to optimize resources.
Road, Civil Aviation, and Railway Sector Analytics with historical trends.
Top 9 Learnings from Building Indian Infra AI Insights:
Scalability through Snowflake: Leveraged Snowflake for handling large datasets, ensuring speed and reliability.
Interactive Visualizations: Used Plotly and Matplotlib to create dynamic and insightful dashboards.
AI Model Integration: Developed and embedded predictive models for accident severity and high-risk area identification.
Real-time Data Processing: Integrated APIs for real-time AQI and traffic updates.
Infrastructure Hypotheses Validation: Validated theories like “better infrastructure leads to regional economic upliftment” using historical and custom datasets.
Streamlit for Seamless Frontend: Streamlit enabled a user-friendly and interactive interface, perfect for exploring insights.
Challenges in Data Cleaning: Navigated through noisy, unstructured datasets and standardized them for accurate analysis.
Focus on Social Impact: Incorporated tools for preventing road accidents, reducing congestion, and improving air quality.
Collaboration and Ideation: Collaboration across data sources like data-gov-in, Snowflake Marketplace, and custom datasets enriched the project.
Explore/Share live app:Indian Infra AI Insights
Demo (~ 15-20 Min) : LinkedIn