yueqqiu.com

Turning vision into systems, stories, and products.





Octo - 即梦Dreamina AI 原生影视创作工具
协作陪伴型创造力Agent - 用户 策略 应用构建
Octo>>是即梦面向影视与视觉创作者推出的 AI Native 创作产品,尝试重新定义创作者与 AI 的关系:AI 不再只是响应指令的工具,而是能共同理解语境、参与创作推进的创意合伙人。作为 0-1 核心成员&创意 Agent 的产品owner,围绕“同屏共感”这一核心体验,定义产品命题、能力结构、交互机制与应用落点、以及可复用的benchmark体系,把多模态能力组织成一套完整落地的创作产品。

产品说明书>>

Octo is an AI-native creative product launched by Bytedance for film and visual creators, exploring a new relationship between creators and AI: not just a tool that responds to prompts, but a creative partner that understands context and actively participates in moving the creative process forward. As a founding core member and product owner for the Creative Agent, I shaped the product proposition, capability structure, interaction model, and application scenarios around the core experience of “shared-screen resonance,” turning multimodal capabilities into a fully realized creative product.




Vibe / Life Coding
把判断快速变成工具和实验,是我近阶段持续在做的一类实践


一手信息雷达
面向融资、IPO 与热点信息获取滞后的问题,我用 Codex 搭建了一套一手信息监控工具,尽量缩短信息从源头到消费端的传递链路,更高效地捕捉值得跟进的市场信号。
First-hand Info Radar
Built a first-hand signal monitoring tool with Codex to track fundraising, IPO, and market-moving information earlier and more efficiently.


创业路径适配测试
为有创业意向的朋友快速搭建了一套 10 分钟交叉测试工具,输出包含产品方向、公司路径、融资或退出方式,以及 90 天行动建议的参考叙事。
GitHub>>
Founder Path Test
Built a lightweight 10-minute diagnostic tool for aspiring founders, generating suggested narratives across product direction, company path, financing or exit route, and a 90-day action plan.




AI 字体特效系统 - 剪映CapCut
AI赋能提效 - 模型数据 工具搭建 效果策略
从 0-1 搭建 AI 字体特效生产系统,探索在成熟业务体系中,AI 如何重构既有产能。作为主负责人全程参与内部模型生产链路搭建,定义可复用的数据样本策略,打磨模型与调参工具,并把控最终效果质量。最终将定制字体模块产能提升xxxx%,且在APP内为此专设“剪映独家”字体tab,获得反响极好的商业化效果。
Built an AI font-effects production system from 0 to 1 within a mature creative ecosystem, leading the model pipeline, reusable data strategy, tuning tools, and output quality. The project increased custom font production capacity by xxxx% and drove strong commercial results, including a dedicated exclusive font tab in the APP.



Emoji AI 贴纸 - 醒图/剪映
AI赋能提效 - 需求洞察 工作流搭建 LoRa
基于图片编辑器应用中,用户“挡脸发图”的使用习惯,以及对emoji和动物形象的偏好洞察,快速搭建 workflow并通过数据策略训练 LoRA 产出高美感 AI 贴纸。一周内完成从洞察到落地,作为成熟应用体系下 AI 快速赋能的轻量案例,比Apple的Genmoji早了两个月。
Based on the insight that users of this photo-editing app often cover their faces in photos and prefer emoji and animal-style expressions, I quickly built the workflow and trained a LoRA model to produce high-aesthetic AI stickers. The project went from insight to launch in one week, serving as a lightweight case of rapid AI enablement in a mature product, two months ahead of Apple’s Genmoji.



战略体系搭建 & 视觉策略 - 初创公司
系统构建 - 业务洞察 战略布局 体系复用
1. 从品牌战略出发,主导双产品体系的命名架构、品牌定位与视觉表达规划,梳理品牌关系与差异化路径,搭建统一且可延展的品牌体系框架。
2. 主导品牌视觉策略与体系建设,基于业务目标和用户洞察,搭建覆盖产品与营销端的可拓展视觉框架;协同产品与研发推进设计规范落地,提升多端产出效率与一致性,并结合 Twitter(X)、Discord、Instagram 等平台特性制定差异化视觉创意方向。
1. Led the naming architecture, brand positioning, and visual expression planning for a dual-product ecosystem from a product strategy perspective, clarifying product relationships and differentiation paths while building a unified and scalable brand framework.
2. Led brand visual strategy and system development based on business goals and user insights, creating a scalable visual framework across product and marketing touchpoints; partnered with product and engineering teams to implement design standards that improved cross-platform efficiency and consistency, while defining differentiated creative directions for platforms such as X, Discord, and Instagram.