---
title: "Shuoxin Wang (@SusieWang) | Hashnode"
description: "Shuoxin Wang (@SusieWang) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@SusieWang"
---

@SusieWang

# Shuoxin Wang

@SusieWang · Joined April 2025

### [Put Your Real-Time Voice Agent on a Skill Roster, Not Under Constant Review](https://trtc.hashnode.dev/put-your-real-time-voice-agent-on-a-skill-roster-not-under-constant-review?utm_source=hashnode&utm_medium=feed)

Use a versioned skill roster to bound voice-agent authority, handle interruptions, route exceptions, and keep every real-time turn observable.

12 min read · [trtc.hashnode.dev](https://trtc.hashnode.dev/put-your-real-time-voice-agent-on-a-skill-roster-not-under-constant-review?utm_source=hashnode&utm_medium=feed)

[2](https://hashnode.com/posts/put-your-real-time-voice-agent-on-a-skill-roster-not-under-constant-review/6aaf42a8900c482fa6e812f7) [Sep 20](https://hashnode.com/posts/put-your-real-time-voice-agent-on-a-skill-roster-not-under-constant-review/6aaf42a8900c482fa6e812f7)

### [Don’t Open Another AI Inbox: Build a Question Queue on Top of Community Chat](https://trtc.hashnode.dev/don-t-open-another-ai-inbox-build-a-question-queue-on-top-of-community-chat?utm_source=hashnode&utm_medium=feed)

Your users are already explaining what they need in group chat. The tempting response is to add an AI bot, a feedback inbox, or a dedicated support room. That creates a second problem: someone must co

11 min read · [trtc.hashnode.dev](https://trtc.hashnode.dev/don-t-open-another-ai-inbox-build-a-question-queue-on-top-of-community-chat?utm_source=hashnode&utm_medium=feed)

[Aug 14](https://hashnode.com/posts/don-t-open-another-ai-inbox-build-a-question-queue-on-top-of-community-chat/6a7e975b7f00be5e08f11dbe)

### [Stream the Preview, Commit the Message: A Reliable Architecture for AI-Assisted Chat](https://trtc.hashnode.dev/stream-the-preview-commit-the-message-a-reliable-architecture-for-ai-assisted-chat?utm_source=hashnode&utm_medium=feed)

Use .NET Minimal APIs, IAsyncEnumerable, SignalR, and a durable social messaging layer without confusing streamed AI output with committed conversatio

13 min read · [trtc.hashnode.dev](https://trtc.hashnode.dev/stream-the-preview-commit-the-message-a-reliable-architecture-for-ai-assisted-chat?utm_source=hashnode&utm_medium=feed)

[Aug 13](https://hashnode.com/posts/stream-the-preview-commit-the-message-a-reliable-architecture-for-ai-assisted-chat/6a7d464c2fdcdf1aa2d3fe6a)

### [Don’t A/B Test the LLM: Randomize the Voice Conversation Policy](https://trtc.hashnode.dev/don-t-a-b-test-the-llm-randomize-the-voice-conversation-policy?utm_source=hashnode&utm_medium=feed)

A production-ready experimentation architecture for AI voice companions that preserves randomization across turn-taking, interruption, latency, and re

12 min read · [trtc.hashnode.dev](https://trtc.hashnode.dev/don-t-a-b-test-the-llm-randomize-the-voice-conversation-policy?utm_source=hashnode&utm_medium=feed)

[Aug 11](https://hashnode.com/posts/don-t-a-b-test-the-llm-randomize-the-voice-conversation-policy/6a7aa31f9942be23a2579bff)

### [Beauty AR Needs a Degradation Receipt, Not Just an FPS Counter](https://trtc.hashnode.dev/beauty-ar-needs-a-degradation-receipt-not-just-an-fps-counter?utm_source=hashnode&utm_medium=feed)

Design measurable, reversible Beauty AR degradation for low-end devices, with user-visible receipts and actionable diagnostics.

11 min read · [trtc.hashnode.dev](https://trtc.hashnode.dev/beauty-ar-needs-a-degradation-receipt-not-just-an-fps-counter?utm_source=hashnode&utm_medium=feed)

[Aug 5](https://hashnode.com/posts/beauty-ar-needs-a-degradation-receipt-not-just-an-fps-counter/6a731e39153b0bdc1e3be181)
