The problem Prabhawatt solves

1. Consumers with solar installations struggle to optimize electricity costs due to complex time-based pricing.
2. No automated way to balance grid vs solar power usage based on dynamic tariffs.
3. Manual scheduling of appliances during low-tariff periods is inefficient & impractical.
4. Limited visibility into real-time energy costs and potential savings.
5. Lack of smart integration between solar production, storage, and consumption patterns.

Target Impact :
Enable users to reduce electricity costs by intelligently managing their energy usage and solar resources through automated, data-driven decisions.

Challenges I ran into

Limited to no data
Since the platform is focused on India, there is little to no information on this as solar adoption is in very early stages in India. I had to simulate a lot of information, like real-time fetching of TOU tariff based on future predictions of feather, peak, and off-peak average rates, etc.

Hallucinations By AI
AI was utilized for forecasting a lot of data produced, but AI being AI led to a lot of misclassification of data, and I had to use libraries like Instructor and Better for prompting for the ideal outputs.

Web3 integration
Being my first time working with web3 technologies, I had a tough time writing smart contracts in solidity. Also, the website is not fully based on web3, it's a mixture of web2 as it uses Firebase authentication. So it was a challenge to bridge the two technologies together.