Skip to content

AI Specialist Corner

Narrow, rule-following agents built to complete specific business tasks reliably.

You don’t need an omnipotent AI agent

The best business agent is not the one that can do everything. It is the narrowest agent that can complete a specific task reliably, under explicit rules and permissions.

Read more →

Prompt Engineering will never be fully deterministic

If you are holding the hope that LLM will one day make the prompt engineering deterministic, you are holding the wrong hope. Prompt engineering is operated in a unnormalized distributed semantics space, it requires less setup, at the cost of less predictability.

Read more →

Table reservation CUI design

The high-level design of a conversational user interface (CUI) for table reservation. It includes detailed descriptions of each functionality, providing a shared understanding among stakeholders.

Read more →

Reuse reservation module to build chatbot

Reuse a pre-defined module to build a table reservation chatbot and manage resources in Google Calendar and Google Admin, allowing for quick integration of reservation functionality into your chatbot.

Read more →

Build a reservation module

Utilize a pre-defined service to build CUI in a module. The step-by-step guide takes table reservation as an example and shows how to customize the service to fit your business needs.

Read more →

From schema to snippets: a blueprint for chatbot

Create powerful custom chatbots with schema-driven development. Define interaction logic by starting from a schema, and then use contextual snippets to describe conversational behavior.

Read more →

Chatbot development with OpenCUI

Revolutionize chatbot development with open-source CUI framework. Streamline the process with a three-layer approach and necessary adaptations for CUI-based nature to deliver business results.

Read more →