Project Management with ChatGPT: Complete Guide

From this lecture you will learn:
• How to communicate during this course

• Where to ask questions

• How to ask questions

• Communication channels

• What is ChatGPT

• What ChatGPT can do

• ChatGPT limitations

• ChatGPT Web Application

• Overview of ChatGPT Web App Interface

• Project Management use cases overview

• Asking simple questions

• Email generation example

• Translation with ChatGPT

• Create Marketing Article with ChatGPT to improve SEO

• Workshop agenda creation example

• Custom Instructions for ChatGPT

• What is a project

• Quality attributes of a project

• Project Charter

• Create a new Project Charter

• Who is a project sponsor

• Total cost of ownership

• Analysis of risks and assumptions in project charter

• Key milestones and planning

• What is a knowledge base

• Overview of tools for knowledge base

• Confluence Cloud

• Advantages of Confluence

• Create knowledge base structure

• Create a Space in Confluence

• What are macros in Confluence

• Review of different macros

• Budget in Project Charter

• Billing Model in Project Charter

• Overview of Time & Material billing model

• How requirements are described and stored

• Feature Specification

• Content of Feature Specification

• Demo of requirements for the project

• Create feature specification with ChatGPT

• Generate a user story with ChatGPT

• BDD format of functional requirements

• Functional VS Non-functional requirements

• The most common non-functional requirements

• Feature specification in the Knowledge Base

• Connect Jira project to Confluence Space

• List user stories in Confluence

• How to change name of the Confluence Space

• Jira Macro Configuration

• Create JQL request with ChatGPT

• Homework

• What is an estimation process and what does it look like

• Planning phase

• Rolling wave planning

• Work Breakdown Structure (WBS)

• Key Characteristics of WBS

• Key Elements of WBS

• WBS Dictionary

• Estimation Techniques

• Expert Judgment

• Analogous Estimation

• Top-Down Estimation

• Story Points Estimation

• Man-Days Estimation

• Three-Point Estimation

• PERT Estimation

• Delphi Method

• Planning Poker

• Bottom-Up Estimation

• Function Points

• T-Shirt Estimation

• Tools for project planning & estimation

• Microsoft Project: Overview

• Advantages & Disadvantages of MS Project

• Other tools for planning

• Custom Estimation Framework review

• Backlog estimation approach

• What is recommended to include in the estimates

• ChatGPT for project planning & estimation

• Plan team structure

• Define risks and opportunities

• Calculate risk treatment expenses

• Calculate risk recovery expenses

• Measurable criteria to evaluate risk probability and risk impact

• Risk management with ChatGPT

• Hourly Rates

• Direct VS Indirect costs

• Calculate communication overhead

• Calculate planned absences

• Calculate management reserves

• Management VS Contingency reserves

• Make data-driven decisions based on the estimation framework

• What is OpenAI API

• What is ChatGPT

• OpenAI VS ChatGPT

• Key Terms and Concepts in OpenAI API

• Prompt

• Tokens

• Models

• Review of key models

• GPT-4 Model

• GPT-3.5 Model

• DALL·E Model

• Whisper Model

• Embeddings

• Moderation Model

• GPT-3 Model

• Point·E Model

• Jukebox Model

• CLIP Model

• Codex Model

• Account creation at OpenAI

• API Reference & Documentation

• Playground Overview

• Manage account settings

• Usage Limits

• Pricing

• Understanding the importance of a "context"

• Configuration of Billing

• Soft & Hard billing limits

• Rate limits

• RPM & TPM

• How to invite members into your organization

• Creation of secret API key

• What is model fine-tuning

• Service-status

• Chat VS Completions API

• When to use Chat API

• When to use Completions API

• Overview of Chat API

• Model Endpoint Compatibility

• Roles: System, User, Assistant, Function

• What “temperature” to use

• Detailed review of Chat API attributes

• model attribute

• messages attribute (incl. role, name, content, function_call)

• Temperature attribute

• top_p attribute

• n attribute

• stream attribute

• max_tokens attribute

• presnece_penalty attribute

• frequency_penalty attribute

• logit_bias attribute

• user attribute

• Authentication & Authorization in OpenAI API

• Selecting model for request

• Send request to OpenAI API GPT model from Postman

• Deprecations

• Parsing GPT response

• “id” attribute

• “object” attribute

• “created” attribute

• “model” attribute

• “usage” attribute

• “choices” attribute

• “message” attribute

• “finish_reason” attribute

• Select programming language

• Overview of official libraries for OpenAI API

• Community libraries for OpenAI API

• Create the first Web Application for ChatGPT Integration

• Review of the Application’s Architecture

• What is function calling

• Business need/Review of use cases

• Function calling algorithm

• OpenAI Chat API review

• function_call: “auto”

• function_call: “prompt”

• function_call: “name”

• function_call: “none”

• Request for a function call in GPT Response

• Code examples review

• Live demo

• Strategy: Write clear instructions

  • Tactic: Include details in your query to get more relevant answers

  • Tactic: Ask the model to adopt a persona

  • Tactic: Use delimiters to clearly indicate distinct parts of the input

  • Tactic: Specify the steps required to complete a task

  • Tactic: Provide examples

  • Tactic: Specify the desired length of the output

• Strategy: Provide reference text

  • Tactic: Instruct the model to answer using a reference text

  • Tactic: Instruct the model to answer with citations from a reference text

• Strategy: Split complex tasks into simpler subtasks

  • Tactic: Use intent classification to identify the most relevant instructions for a user query

  • Tactic: For dialogue applications that require very long conversations, summarize or filter previous dialogue

  • Tactic: Summarize long documents piecewise and construct a full summary recursively

• Strategy: Give GPT time to "think"

  • Tactic: Instruct the model to work out its own solution before rushing to a conclusion

  • Tactic: Use inner monologue or a sequence of queries to hide the model's reasoning process

  • Tactic: Ask the model if it missed anything on previous passes

• Strategy: Use external tools

  • Tactic: Use embeddings-based search to implement efficient knowledge retrieval

  • Tactic: Use code execution to perform more accurate calculations or call external APIs

  • Tactic: Give the model access to specific functions

• Strategy: Test changes systematically

  • Tactic: Evaluate model outputs with reference to gold-standard answers

• TCP/IP/Port

• Static and Dynamic IP

• Port forwarding

• How to configure port forwarding on your router

• Troubleshooting hints

• What is slack

• Slack Installation

• Create an account in Slack

• Create a workspace in Slack

• Create a channel in Slack

• Create an Application in Slack

• Creation of app from scratch

• Creation of app from an app manifest

• Incoming and Outgoing Webhooks

• Events API in Slack

• Configure outgoing and incoming webhooks in Slack

• Architecture overview

• Review of code examples

• How to add the app to a channel

• Slack event payload review

• Slack Java SDK

• Overview of SDK for other programming languages

• Scopes in Slack

• Configuration of required scopes for the app

• Set up app picture

• How to manage the context during the integration with GPT

• Code examples review

• How to limit the context length

• How to manage the context of the Slack Team

• Remove all messages from a channel

• What is Jira

• Jira Editions & Deployment Options

• Advantages of Jira Cloud

• Jira analogs & competitors

• Create an account in Atlassian

• Create Jira Project

• Kanban Template VS Scrum Template

• Jira Kanban Board overview

• Jira project settings overview

• How to add issue types to the project

• How to add fields to issue type

• Create fake data in the project

• Jira API Review

• Jira API Versions

• Authentication & Authorization in Jira API

• What is Forge App in Atlassian

• What is Connect App in Atlassian

• JWT

• 3LO

• Create API Token for Atlassian Account

• Encrypt credentials in Base 64 encoding

• Jira API calls via Postman

• Jira API calls from web app

• Jira API documentation overview

• How to integrate Jira API with GPT, Slack and Web Application

• Connect Jira as a separate datasource

• Read the required context from the Jira

• Function calling implementation to fetch required Jira items

• Demo of real-life application

• Web Application architecture overview

• Best practices

• Generate work item description with the help of ChatGPT

• Create work item in Jira using chat interface

• Assign team member for Jira ticket directly from slack

• Set the due date for Jira work item from Slack

• Generate an email

• Send an email using email address using chat interface

• Send an email using the name of the person using chat interface

In the video I’m going to hold a demo of few more use cases of how you can use our bot developed during the course together with ChatGPT to manage Scrum Team, calculate average velocity and plan a new Sprint. Also in the video I’m going to show you Jira board configured for execution of Risk management and how I use my bot to manage risks on my projects.

In the lesson I used the approach with function callback that we learned in previous lessons.

You can find source code in attachments to the video.

In case you will have any questions, please, post your questions below the video and I will be happy to answer.

During the video we will review:

• How to calculate average velocity with ChatGPT

• How to plan Sprint based on priorities and team’s capacity

• How to check status of risks

• How to work with not processed risks

• How to execute risk management operations using chat interface

• What is fine-tuning

• Fine-tuning steps/algorithm

• Labeled data

• Few-shot learning

• Meta-learning or learning to learn

• Model-Agnostic Meta-Learning (MAML)

• Reptile

• Prototypical Networks

• Matching Networks

• Memory-Augmented Neural Networks

• Transfer learning

• Difference between meta-learning & transfer learning

• Advantages & benefits of fine-tuning

• Use cases and examples when we need fine-tuning

• Available models for fine-tuning

• Chat API VS Completions API

• Ada, Babbage, Curie, Davinci models overview

• Fine-tuning costs

In this lesson, I’m going to share with you important notes about the most recent updates before you start watching the next lessons about fine-tuning.

• Training dataset format

• What is JSONL

• How much examples we need for training dataset

• Data Augmentation

• Overfitting

• Availability & Feasibility of Data

• Classification use cases and examples

• Conditional generation use cases and examples

• Guidelines for preparing training dataset for classification use cases

• Guidelines for preparing training dataset for conditional generation use cases

• What is learning rate

• What is epoch in machine learning

• Requirements for our first custom chat bot

• Prepare training dataset using ChatGPT

• Prepare dataset for validation

• OpenAI Python Client

• Install Python

• What is PIP

• Install OpenAI Python Client

• Prepare datasets for training

• Fine-tune the model

• What is the batch size

• What is the loss weight

• Demo of using of fine-tuned model

• Analysis of fin-tuned model

• How to analyze training process

• Practical Use Case Demo: Chat Bot for Knowledge Base

• Integration of custom chat bot via Slack Messenger

• How to build iterative process of fine-tuning

• What is a metric

• Examples of metrics

• When to use metrics

• What is a KPI

• Examples of KPI

• When to use KPI

• Metric VS KPI

• What is OKR

• Examples of OKR

• When to use OKR

• KPI VS OKR

• What is RAG Status

• RAG status to present KPI

• Why we use RAG status for KPI

• What we are going to learn in this section

• Why this section is important

• Overview of Engineering Excellence Metrics Library

• Web Development related examples and use cases

• Tech Debt Ratio

• Tech Debt Index

• Cyclomatic complexity.

• Definition

• Use cases

• How to measure

• How to read values and what do they mean

• Recommended KPIs

• Recommended Actions

• Unit Testing

• How Unit Tests Work

• Benefits of Unit Tests

• Challenges and Limitations of Unit Tests

• Unit Test Run Success Rate

• Unit Test Code Coverage

• Incremental Unit Test Coverage

• Duplicate Code

• Duplicated Lines

• Duplicated Blocks

• Duplicated Files

• Density of Duplicated Lines

• Commented Code Index

• What is a Code Review

• Code Review Feedback Loop Time

• Code Reviews Amount

• Rules Compliance Index (RCI)

• Violations

• Differences between RCI and Violations

• What is Integration Testing

• What is End-to-End Testing

• Integration VS End-to-End Testing

• Integration Test Coverage

• End-to-End Test Coverage

• Introduction to ChatGPT

• What is ChatGPT

• Brief history of development of ChatGPT

• ChatGPT Capabilities

• Applications in Customer Service

• How ChatGPT can assist in customer interactions

• Examples of ChatGPT in real-world customer service scenarios

• AI in Sales

• How AI tools like ChatGPT can boost sales efforts

• Case studies of AI-driven sales success

• Enhancing customer success with ChatGPT

• Benefits of using AI for customer retention and satisfaction

• How ChatGPT can streamline customer inquiries and provide automated responses

• Introduction to advanced functionalities of ChatGPT beyond basic usage

• Understanding how ChatGPT can be tailored for specific customer service needs

• Strategies for maximizing efficiency in handling customer inquiries using ChatGPT

• Best practices for seamless integration of ChatGPT into diverse customer service workflows

• Discussion on overcoming challenges and potential pitfalls in implementing ChatGPT in customer service

• Analyzing case studies of innovative and successful ChatGPT implementations in various industries

• E-commerce Case Studies

• Implementation process of ChatGPT in E-commerce

• Results and Improvements Observed

• Healthcare Case Studies

• Implementation process of ChatGPT in healthcare

• Results and improvements observed

• Finance Case Studies

• Implementation process of ChatGPT in finance

• Results and improvements observed

• Extracting key insights and lessons learned from real-world examples of ChatGPT in customer service

• Enhancing Customer Interactions with ChatGPT

  • Benefits of using ChatGPT for customer service

  • Examples of improved customer experiences

• Handling Common Inquiries Effectively Using ChatGPT

  • Identifying frequent customer questions

  • Crafting effective responses

• Implementing Escalation Procedures for Complex Customer Issues

  • Recognizing when to escalate an issue

  • Steps for proper escalation

• Importance of Chatbot Etiquette and Maintaining a Human-Like Interaction

  • Best practices for polite and friendly responses

  • Techniques to make interactions feel natural

• Explanation of role-playing learning concept

• Roles & Responsibilities (customer, customer service representative, observer)

• Briefing on different scenarios

• Scenario 1: Handling a customer complaint

• How to generate the most common customer requests

• Prepare instructions for ChatGPT for handling customer requests

• Homework Review:

  • Scenario 2: Assisting a customer with a product issue

  • Scenario 3: Answering a customer query about a service

• Strategies for Maintaining a Positive User Experience with ChatGPT

  • Tailoring responses to match user inquiries effectively.

  • Implementing conversational design principles for a seamless interaction.

  • Ensuring consistency in tone and language across conversations.

• Utilizing ChatGPT Analytics for Improved Response Accuracy

  • Introduction to ChatGPT analytics tools for performance evaluation.

  • Analyzing response accuracy metrics to identify areas for improvement.

  • Leveraging analytics insights to refine ChatGPT's response generation algorithms.

• Enhancing Customer Satisfaction through ChatGPT

  • Understanding the correlation between response accuracy and customer satisfaction.

  • Strategies for proactively addressing user queries and concerns.

  • Personalizing interactions to enhance user engagement and satisfaction.

• Continuous Improvement and Adaptation based on Customer Feedback

  • Importance of requesting and analyzing customer feedback on ChatGPT interactions.

  • Implementing feedback-driven improvements to enhance ChatGPT's performance.

  • Agile adaptation of ChatGPT based on evolving user preferences and needs.

• Measurement and Evaluation

  • Defining KPIs for measuring the effectiveness of ChatGPT in customer experience.

  • Establishing benchmarks for response accuracy, user satisfaction, and engagement.

  • Incorporating feedback loops for continuous monitoring and improvement.

• What is Chatbase

• Chatbase Features

• Chatbase Pros and Cons

• What is better: Chatbase VS Custom-made Chatbot

• How Chatbase works on the backend

• Chatbase pricing plans

• How to create Team in Chatbase

• Create Chatbot

• Train Chtabot

• How to select training data sources

• Chatbot settings and configurations

• UI customization of the chatbot

• Chatbot for Revolut Bank demo

• How to analyze chatbot activity

• Filtering of chatbot interactions

• Team settings

• Introduction: What is Revolut

• Customer Service at Revolut

• Help Center overview of Revolut

• How to gather data for training ChatBot for Revolut Support

• Overview of ChatBot Infrastructure

• ChatBot Demo: How ChatBot may handle the most common customer requests

• Tools for API testing: Overview

• SOAP

• Comparative analysis of Postman, Fiddler and Wireshark

• Postman

• Key components of Postman

• Postman Desktop VS Web Version

• Postman installation

• Postman Interface Review

• Import/Export collections

• Create new collections

• Sending requests with different HTTP methods

• Reading response in Postman

• Sending and reading custom headers

• Testing of RESTful service

• CRUD Operations with resource using REST API

• Difference between form-data and x-www-form-urlencoded