Deep-Dive into AI Use-Cases¶
In this module, we will undertake a comprehensive exploration of Project Miyagi and the Reddog codebase. This deep dive will highlight a range of use-cases that can be adopted into your own applications, thereby enabling you to infuse AI to address the diverse needs of your end users.
For detailed implementation of the use-cases, see miyagi/services
Use Cases¶
- Synthesis: Learn how information is synthesized in Miyagi to generate insights, orchestrated by Semantic Kernel.
- Multi-modal Generation: Discover how Generative AI is used in Miyagi and Reddog to generate text, images, and videos.
- Conversation: Deep-dive into how Miyagi Chatbot is built using Semantic Kernel's Copilot Chat, which you can leverage to build your own Copilot.
- Summarization: Understand how to create TL/DRs in your apps using Miyagi's example.
- Translation: Language translation.
- Code generation: Code generation with SK planner and CoT.
- Classification: Explore how Miyagi uses SK to classify expense categories.
- Speech-to-Text: Leverage Whisper to perform SoTA speech-to-text, which is used in Miyagi chatbot.
- Semantic/Neural Search: Learn how to improve search results with AI's understanding of context.
- Anomaly Detection: Discover how AI can identify unusual patterns that could indicate potential issues.
- Plugins: Understand how to extend and enhance your applications with ChatGPT and BingChat plugins.
- Agency and Planning: Learn how Miyagi uses SK's planner and Jarvis for agentic planning to rebalace portfolio based on user's preferences.