Week #4 #

External Feedback #

Feedback Collection #

This week, we received comments from the client, provided by experts. The feedback covered various aspects of our solution, including the use of technologies, algorithms, system errors, and interface improvements.

Feedback Analysis #

Key points from the client’s feedback:

1. The use of OpenAI ChatGPT for speech recognition needs to be adjusted according to the requirements.
2. The algorithm for how the bot recognizes and implements changes in Federal Law 44 needs to be described.
3. An error occurred during authorization with the provided data.
4. The fields in the purchase request editing are not translated into a readable format.
5. The “finish editing” button does not work when editing a purchase request.
6. The logic for selecting products for forecasting needs to be refined to ensure usability when there are many products.

Actions Taken Based on Feedback #

1. We replaced the use of ChatGPT with BERT for speech recognition to meet the requirements.
2. We started redesigning the bot’s algorithm for recognizing and implementing changes in Federal Law 44.
3. We fixed authorization errors.
4. We improved the readability of fields when editing purchase requests by adding necessary translations.
5. We checked the functionality of the “finish editing” button and fixed the bug.
6. We began working on improving the product selection logic for forecasting to ensure usability when there are many products.

Testing #

Testing Process #

• Conducted functional testing to identify and fix errors mentioned by the client.
• Conducted user interface testing to improve the readability of fields and check the functionality of buttons.

Testing Results #

• Identified and fixed the authorization error.
• Found that the fields in the purchase request editing are not translated into a readable format.
• Identified that the “finish editing” button does not work.

Next Steps in Testing #

• Continue testing the product selection logic for forecasting.
• Conduct additional tests to verify the functionality of algorithms and interface after changes.

Iterations and Refinements #

Iteration Process #

• After receiving the feedback, we immediately started making changes and corrections to the project.
• The main focus was on adjusting functionality and improving the user interface.

Implemented Changes and Improvements #

• Replaced ChatGPT with BERT for speech recognition.
• Fixed authorization errors.
• Updated the interface design.
• Enhanced the feature for tracking the stock of selected products.

Plans for Future Iterations #

• Continue working on the client’s suggested revisions, including improving the product selection logic and describing the bot’s algorithms.
• Improve the readability of fields when editing purchase requests.
• Fix the “finish editing” button functionality.

Progress Documentation #

Main Achievements of the Week #

• Replaced ChatGPT with BERT.
• Fixed authorization errors.
• Updated the interface design.
• Enhanced the feature for tracking product stock.

Challenges and Problems #

• The need for significant changes in the bot’s algorithms.
• Issues with the readability of the interface when editing purchase requests.
• The error in the “finish editing” button functionality.

Plans for the Next Week #

• Continue working on the client’s revisions.
• Improve the product selection logic for forecasting.
• Refine and test the purchase request editing functionality.