Recap for October 9th, 2024 - Monthly Semantic Kernel Office Hours (APAC)

[sophialagerkranspandey]

Monthly Semantic Kernel Office Hours for APAC - October 9th, 2024 recording: https://www.youtube.com/watch?v=32ePRMcaksI&list=PLmsFUfdnGr3xAtM3ofjpEkIWnuSA2d-i-&index=36

If you weren't able to attend office hours but still have questions, please post them here: https://github.com/microsoft/semantic-kernel/discussions

Meeting notes:

• Software Releases: Evan announced recent software releases, including updates for .NET and Python, emphasizing the team's continuous improvements and the availability of these updates on their repository. 1:09
• Documentation Enhancements: Evan highlighted significant enhancements in documentation, including updates to the Python and C# API references, aiming to make it easier for users to find information and utilize the software effectively. 1:37
• Process Framework: Evan introduced a new process framework, detailing its capabilities for event-driven and stateful AI processes, which can be auditable and include telemetry. This framework is designed to streamline business processes using AI. 4:54
• Customer Use Case: Vic presented a customer use case from preezie, showcasing an AI Shopping Assistant designed to improve online shopping experiences by providing sizing recommendations and product bundling suggestions. 11:00
• Developer Experience: Dylan McCarthy inquired about enhancing the developer experience with semantic kernel, specifically through demos or examples that leverage GitHub models for ease of development and testing. Evan responded positively, suggesting the use of notebooks for quick starts. 23:36
• Production Hosting: Luke raised a question about the best way to host semantic kernel for production, with Evan suggesting app services for large enterprises or Azure functions for local testing, highlighting the flexibility in deployment options. 39:00
• AI Chat Agents: Dylan McCarthy discussed using semantic kernel to create AI chat agents for live chat streams, aiming to intelligently respond to relevant queries while ignoring others, with Evan providing insights on function calling and system prompts to guide agent behavior. 32:46