Build 5 non-trivial AI apps in 5 weeks

Build AI Chatbot
– Deep Dive

• Set up a development environment
  Have a solid foundation in setting up development environments tailored for AI chatbot creation.
• Leverage AI SDKs
  Be skilled in integrating various AI SDKs for effective chatbot functionality across multiple platforms.
• Design Integration Flows
  Demonstrate the ability to design and implement efficient code review integration flows within Discord and GitHub.
• Interface with AI Models
  Gain expertise in interfacing with multiple AI models using LangChain.js for comprehensive code analysis.
• Facilitate AI-Driven Feedback
  Learn to automate sending code for AI review and integrate feedback directly within communication platforms.
• Construct Advanced Prompts
  Develop the skill to create complex system prompts for detailed AI feedback, enhancing code review quality.
• Utilize Embeddings for Search
  ‍Understand how to implement search functionalities using embeddings, improving data retrieval based on past interactions.
• Customize AI for Application Needs
  Acquire the ability to customize and extend AI models, particularly Codex AI, for specific application features, elevating the bot's utility and integration within development workflows.

• Set up and configure development environments using Next.js and Deno for AI chatbot projects.
• Use AI SDKs for seamless integration with platforms like Discord and GitHub.
• Design and execute an integration flow for direct code reviews within Discord and GitHub pull requests.
• Utilize LangChain.js as a unified interface for interacting with various AI models for code analysis.
• Implement feedback loops, sending code to AI services and posting reviews directly on Discord.
• Develop a system for constructing detailed prompts, potentially outputting as JSON for precise feedback.
• Create a search mechanism based on bot interactions, utilizing embeddings for efficient data retrieval.
• Explore the customization and enhancement of AI models, specifically Codex AI, for tailored application features.

Build Your Custom LLM

• Understand LLM Fundamentals
  Acquire a solid understanding of LLM architecture and its capabilities in processing and translating natural language.
• Build Custom LLMs
  Learn to build a custom LLM tailored to specific tasks, like SQL query generation, from the ground up.
• Deploy Advanced Models
  Gain practical experience in locally setting up and deploying advanced models like LLama2-70B and Mistral-8x7B.
• Utilize Hugging Face Effectively
  ‍Master using Hugging Face for model deployment, dataset management, and leveraging community spaces.
• Implement NVIDIA Technologies
  Understand how to set up and use NVIDIA Chat RTX for real-time model interactions and enhancements.
• Create a Query Chatbot
  Develop a chatbot that can process natural language queries and extract specific information from large databases, demonstrating a practical application of LLMs in data retrieval and management.

• Develop a foundational understanding of LLMs and their application in translating natural language to code.
• Create a basic LLM from scratch using Python.
• For local use, set up and utilize advanced LLMs like LLama2-70B and Mistral-8x7B.
• Explore Hugging Face's ecosystem for model deployment and dataset management.
• Use NVIDIA Chat RTX to create a model for local interaction.
• Specialize in a custom LLM for SQL query generation.
• Deploy a functional chatbot capable of querying a large database.

Visual AI – Storytelling App

• Master Visual AI Tools
  Gain proficiency in leveraging Midjourney and Dall-E to craft engaging, thematic images tailored to storytelling needs.
• Enhance Stories with Audio AI
  Learn to utilize audio AI technologies like ElevenLabs and PlayHT to bring stories to life through realistic speech and sound effects.
• Create Consistent Visual Narratives
  Develop the skill to generate a coherent series of images that visually narrate a story, enhancing the viewer's experience.
• Integrate Speech-to-Text for Transcription
  Master speech-to-text APIs for accurate transcription, facilitating seamless integration of narrative content.
• Narrate with Text-to-Speech
  Acquire the ability to convert text into natural-sounding speech, using tools such as ElevenLabs to narrate stories effectively.
• Produce Narrated Video Slideshows
  Learn to compile images and narrations into a captivating video slideshow, completing the storytelling experience from start to finish.

• Master using visual AI tools (Midjourney, Dall-E) to generate compelling images for storytelling.
• Utilize audio AI technologies (ElevenLabs, PlayHT) to enhance storytelling.
• Generate consistent and thematic images for storytelling.
• Implement speech-to-text APIs for accurate transcription.
• Craft compelling prompts for AI-driven image creation.
• Synthesize speech from text to narrate stories.
• Compile a series of images and narrations into a video slideshow.

Vector DBs for knowledge retrieval

• Understand Vector Databases
  Gain in-depth knowledge of the workings of vector databases and their role in AI for efficient data management and search.
• Acquire Advanced AI Integration Skills
  Learn to integrate Chroma, Picone, Quadrant, and OpenAI vector databases into applications, enhancing problem-solving capabilities.
• Develop AI Plugins for Zoom
  Master the design and implementation of AI plugins for Zoom, enabling automatic transcription, meeting summarization, and searchable knowledge base integration.
• Implement Searchable Knowledge Bases
  Learn how to create knowledge bases using vector-based search technologies, including pg_vector in PostgreSQL, vector-vss in SQLite, and Typesense for on-the-fly embeddings.
• Integrate with Corporate Wikis
  Understand how to connect AI-enhanced search with wiki systems like Confluence or Notion, improving knowledge management and accessibility within organizations.
• Apply AI in Organizational Contexts
  Demonstrate the application of AI to streamline meeting note management and insights retrieval, showcasing AI's impact on workplace efficiency.

• Understand the role of vector databases in AI and knowledge management.
• Explore the capabilities of Chroma, Picone, and Quadrant for AI applications.
• Learn to implement pg_vector in PostgreSQL and vector-vss in SQLite for vector-based search.
• Utilize Typesense to enhance search functionality with on-the-fly embeddings.
• Develop an AI Zoom plugin that automatically transcribes audio, summarizes, and integrates meeting notes into a searchable knowledge base.
• Integrate vector-based search with wiki systems for efficient retrieval of past meeting insights.

Information aggregation for custom GPTs

• Master Content Aggregation
  Gain proficiency in aggregating relevant information from diverse web sources for data-driven applications.
• Automate Data Ingestion
  Learn to automate the process of data ingestion and updates, ensuring the AI model remains current with the latest information.
• Develop Custom GPT Models
  Acquire the skills to develop and maintain custom GPT models tailored for specific needs or topics, like new hire onboarding.
• Implement Dynamic Information Pipelines
  Understand how to build and manage dynamic pipelines that feed real-time data into AI models for up-to-date outputs.
• Apply Custom GPTs in Real-world Projects
  Demonstrate the practical application of custom GPTs in projects, showcasing their ability to transform and personalize user experiences.

• Learn to build web scrapers for content aggregation from multiple sources.
• Utilize tools like youtube-dl for downloading video transcriptions.
• Set up queues and Cron systems for automated data ingestion and updates.
• Construct a data pipeline to keep custom GPT models updated with fresh information.
• Understand the fundamentals of OpenAI's custom GPTs, including creation and application.