Build your own Chatbot using Python, ChatGPT & WhatsApp

Sign up on OpenAI to access the API, create credentials, and get started with ChatGPT integration by entering your email, password, and agreeing to terms.

Retrieve your OpenAI API key from openai.com, create a new secret key, and copy it to a notepad for later use since the site won't remember it.

Explore ChatGPT's playground to chat with the model in a conversational context and test prompts, learning to adjust options such as temperature and max length for tailored responses.

Explore ChatGPT API parameters, including model selection, prompt, length, temperature, top_p, and stop sequences, and learn about the DaVinci models like text-davinci-003 and their multilingual capabilities.

Learn how to handle error code 429 during spreadsheet tasks in this course, by deleting the cell and re-dragging the equation, noting ChatGPT demand.

Build a Google script that enables a Google spreadsheet to query ChatGPT via OpenAI API and return the answer, using a secret key, model text-davinci-003, temperature, and max tokens.

Explore deeper explanation of AskChatGPT function in .gs by fetching a URL with API compliant options, parsing JSON, selecting the highest confidence first result, and trimming noise from text.

Automate ChatGPT queries in Google Sheets using a script and concatenate to create multiple country questions and fetch the most confident answers alongside GDP per capita.

Regularize ChatGPT's answer formatting by applying financial formatting to numbers, standardizing currency symbols and removing unnecessary decimals for consistent, clean outputs.

Learn how to ask ChatGPT for information sources and dates, verify data with cited references, and cross-check numbers like GDP per capita against reliable sources.

Learn to force a specific output format from ChatGPT, extracting amount, year, and reference links, then automate with Google Sheets and verify sources across countries like Spain and the UK.

Plot the core relationship between energy consumption per capita and GDP per capita using ChatGPT-provided data, visualize with charts, and double-check sources for accuracy.

Wraps up by showing how to plot life expectancy against GDP per capita with a last reported life expectancy column, and how to request data from ChatGPT for clean plotting.

Develop a categorization method with ChatGPT natural language processing, call the categorize function, define your own categories, and test that the approach works for every country on the list.

Learn to build a google script function that takes a list as an argument and uses ChatGPT to categorize input into predefined categories within a Node.js workflow.

Show how to build chatbot tools using Microsoft Visual Basic and Excel, replacing Google scripts, and introduce the ask ChatGPT method to get answers like what's the capital of Italy.

download and install the diva-json module, import it into Excel's Visual Basic editor, and enable Microsoft XML 6.0 and Microsoft Scripting Runtime references to connect to external APIs.

Build a VB macro that posts a JSON prompt to the OpenAI API via MSXML2.ServerXMLHTTP, sets authorization and content-type headers, parses the JSON response, and returns the first choice text.

Practice building a Visual Basic module that constructs a question from categories via a for-each loop and uses ChatGPT to classify the subject.

We present the project outcome: building a Python chatbot with the ChatGPT API and OpenAI library to sustain contextual conversations. It shows how context guides responses.

Use a reserved phrase like 'close chat' with ChatGPT to break a Python while loop, enabling clean termination of a conversation while planning to manage memory and context.

Implement a conversation history to give your chatbot memory by storing user prompts and ChatGPT responses, updating history after each exchange, and sending the full context with every query.

Learn to save and load conversational context for a chat bot using txt files, with conversation IDs, history management, and load/save methods.

This module presents a project that builds an automated chatbot using Python, Flask, WhatsApp, Twilio, and ChatGPT to answer questions via WhatsApp.

Explain how the whatsapp chatbot architecture routes messages from a Twilio number to a local Flask server, exposes it publicly with Anjirak, queries ChatGPT, and returns replies via Twilio.

Set up and run a Flask server with two routes: a home route that confirms the server is up, and a Twilio receive message route that returns a message.

Test the Flask app's receive_message route via ngrok by hitting the /twilio/receive_message endpoint to confirm that a message is received.

Configure Twilio WhatsApp sandbox to send messages to your Flask app via a public Ngrok URL, then handle post requests and parse the message body to power your chatbot.

Post Twilio WhatsApp messages to ChatGPT via the OpenAI library in a Flask app, using a helper OpenAI API module to manage conversations and generate responses.

Assign each client a private conversation by using the sender id from the request form as the conversation id, cleaning with re.sub, and saving exchanges with conversation.save_conversation.

Demonstrates storing the conversational context in a txt file, clearing the WhatsApp chat, restarting the Flask server, and ensuring ChatGPT answers with the saved favorite movie John Wick.

Celebrate completing the course and stay ahead as you harness natural language processing and GPT in your own custom-made applications and software, continuing your journey into this revolutionary technology.