---
title: "Unique Friend | Tyrion.AI’s Liu Yuchen: Using Foundation Models Well Means Thinking AI-Native About Enterprise Software"
author: "Unique Research"
sourcePublication: "Unique Research (FFCAP English library)"
originalPublishedAt: "2024-01-21T14:00:00+08:00"
canonical: "https://ffcap.cn/en/research/20240121a03s7c00"
source: "https://ffcap.cn/en/research/20240121a03s7c00"
language: "en"
cnSourceUrl: "https://view.inews.qq.com/a/20240121A03S7C00"
---

# Unique Friend | Tyrion.AI’s Liu Yuchen: Using Foundation Models Well Means Thinking AI-Native About Enterprise Software

_Unique Research · Insight / Unique Friend Q&A · January 21, 2024 · Guest: Liu Yuchen, founder & CEO, Tyrion.AI (意言科技)_

Inventory type remains **insight** (Unique Friend series — not partner welcome).

## Guest

**Liu Yuchen** — founder & CEO, Tyrion.AI (意言科技). Former CDO, L’Oréal PPD China (digital transformation); formerly led industry AI SaaS products at Tencent Cloud (0→1); earlier industry pre-sales lead at Neustar in the US on data-middle-platform builds for Citibank, TD Ameritrade, Capital One, Fidelity, Amica Mutual and peers (Forrester case win). Prior: Accenture, Ogilvy.

## Q&A

**1\. Product and core scenes?**  
Core product is the enterprise Agent build-and-ops platform **Tyrion**. Using foundation models, it automates business processes and captures expert knowledge — Agents as “digital employees” that co-work with humans on complex, repetitive tasks while enterprises deposit best practices and keep optimizing.

**2\. R&D / GTM — 2023 challenge and win?**  
Tyrion lets firms build AI Agent workflows that lift headcount efficiency across departments. No-code design lowers the bar to create and use Agents. It supports mainstream foundation models, light fine-tuning for small scenes, and for security-sensitive large enterprises: on-prem model deployment plus edge AI hardware. Despite a short history, Tyrion has signed Fortune 500 manufacturers, pharma, finance, and FMCG brands, and partners with model, cloud, and compute vendors. Biggest 2023 challenge: finding the intersection of model capability and real business need — what can / cannot / should / should wait. Biggest win: customer feedback that deepened vertical-scene depth.

**3\. Opportunities vs. past waves?**  
LLMs and Agents rewrite enterprise software: interaction that understands human intent, and far more flexible task judgment and handling than classical software. That unlocks latent enterprise demand and gives AI-Native companies a shot at Salesforce- or SAP-scale giants. Many ToB vendors shipped “AI features” last year as bolt-ons to existing products. To truly use GenAI’s energy, rethink enterprise software _AI-Native_ — look again at scattered, once-human-only flows that machines can now take. Huge opportunity there.

**4\. Human–machine collaboration?**  
Employees and enterprises both benefit. For employees: baseline white-collar work shifts to Agents, freeing time for creative and planning work; strong operators can move to new business and harder tasks as Agent process control improves. For management: information and process become more efficient; energy moves from low-efficiency people management to running the business; best practices deposit and iterate automatically — the AI-era ask for efficiency and innovation.

_1 CN image; attach on ingest. Signup CTA in CN omitted._

— END —

---

Editorial English edition published on UniqueCapital: https://ffcap.cn/en/research/20240121a03s7c00
On-site reading page: https://ffcap.cn/en/research/20240121a03s7c00