The IoT-Partner chatbot is an intelligent assistant designed to assist users with smart agriculture and IoT-related queries. The chatbot provides real-time support for topics such as IoT sensors (temperature, humidity, soil moisture), devices like ESP32 and Raspberry Pi, and network configuration. It aims to help both farmers and engineers by offering practical solutions and advice, regardless of their experience level.
This chatbot was developed as part of a competition and is inspired by the challenges faced in the development of a Smart Agriculture with IoT project, which involves the integration of hardware and software using sensors, ESP32, and other IoT technologies.
The chatbot's backend leverages the Groq API to generate responses based on the input from users. Users can ask questions related to IoT devices, sensors, and troubleshooting. The system is designed to:
1.Provide responses using a language model optimized for agricultural IoT.
2.Enable real-time chat capabilities through the Streamlit framework.
3.Support personalized recommendations for users in smart farming contexts.
The system also offers troubleshooting advice to resolve common issues with devices like ESP32, sensors, or network configurations.
.png?Expires=1738880915&Key-Pair-Id=K2V2TN6YBJQHTG&Signature=lGwNmXL~Znm1T4aelLwrEgoxV3RShQSNdL6AnvrRG4HpFuSKKxO6yXk4M2YQV9OF1NHZ962YPo~N2uJAxQsY6j6PO6vup2cAUaKkEY8gLEofCSlhRT9fGhbC05Qx9ngIjYZ8bWpxkguDp0lGPMUaaGgmOVyEwlM-IaglpZqc6dS5dJe5u44iR7artum7s7TWQ1lJrXArocIIWQAy~AENiHilUGOrkB6EEC7O85Y-uPSHeZSN1CjOsdKRYDupYzKBlNLddKl9MWGtZ8pEJjUVOd5pItTX5LsXBpdE8z2ZC1dZvVBGNr6RFOLHQ6ZcWyx6YHYSpEWhlZT90wpiWWYcfA__)
.png?Expires=1738880915&Key-Pair-Id=K2V2TN6YBJQHTG&Signature=BHYwYodLVsk4Kk~DUl4~xb~vKjOZvIFxeiXK4haeXARTaIXYdDoe9juECcDnNoEpZUdq33vBt5t0-qGwgSQg21NDtE-XUOz9F0rf6~gb6MSlPqc3LuzcW9npyTOxwEEgajDm54t2ESesuiErTdre4~DnfJE-9o-iAauFE4~~HT-5haihy7D4FhUP3CIyp0rlyutKGRLSBRJ~bxxXLhxooiq-Hw9zw7mdoZ-WO0oueFKWUOggAzcDWv17xu~~pf32OPovNFaX09uqzscl-sXKsEoXuC4j1Rl8uaz5ez9~R0iqSdgQ3nBIQIwSPV7YEmK5XNAAT7J6uksfwBxhsulHPA__)
One of the major challenges in this project was integrating the Groq API into the chatbot to provide natural language responses. Initially, API responses were inconsistent, and latency issues affected the chatbot’s real-time performance. This was resolved by adjusting API settings and improving the connection configuration.
The Streamlit framework was chosen for its ease of use, but customizing the chat interface in Streamlit posed some challenges. Initially, the interface did not render properly in dark mode, and there were layout issues. After troubleshooting, I was able to create a more responsive design that is visually appealing in both light and dark modes.
A key challenge was handling the real-time streaming of chat responses. The bot needed to output text word by word to mimic natural conversation. Implementing a delay between each word and ensuring that the responses flowed naturally required careful timing and handling of the API response.
During development, several connectivity issues occurred while trying to clone the GitHub repository, including SSL/TLS handshake errors and proxy-related issues. I had to troubleshoot the Git configuration and network settings to resolve these problems before being able to successfully clone the repository and push changes.
Since the chatbot was intended for a niche audience (smart agriculture), there was limited training data available for the model. I faced challenges in getting the model to respond effectively to all user queries, particularly for more specific IoT topics. This was mitigated by creating a well-defined system prompt that gave the chatbot clear instructions on its purpose.
There are no models linked
There are no datasets linked
There are no datasets linked
There are no models linked
