WANGENIUS

I launched CMOCHAT in 2025 as a global marketing resource matching platform. I lead product strategy, the data model, and the AI matching engine, structuring the scattered data of CMOs, agencies, channels, and tools worldwide while building a conversational growth cockpit that turns any business goal into an executable go-to-market playbook. I also designed the lead scoring system, incentive mechanics, and settlement tooling so a founding team can move from brief to closed transaction within a week.

I founded Genesis Cosmos with partners and launched Jezzlab, an AI-native creative platform. I orchestrated the product blueprint, interaction language, and agent workflows, introducing concepts such as a worldview operating system and character-generation rails so creators could turn narrative assets into collaborative, systemized tools. We adopted a local-first architecture to protect privacy, layered in visual prompt debugging, multi-writer storyline versioning, and subscription commercialization, and secured the first wave of high-paying users with a tiny team.

I spent a year embedded in Gaorong Capital's AI research team, dissecting tracks such as AI coding, enterprise applications, and Agent Coin from an investor's perspective. I built data warehouses, mapped entire ecosystems, wrote industry reports, and partnered with portfolio founders on KPI dashboards. I conducted in-depth founder interviews, evaluated technical roadmaps, and used model benchmarks plus case reviews to estimate product ceilings, then distilled the findings into internal research baselines and due diligence checklists.

Since 2023 I have worked full time as an independent developer, building a solo pipeline that runs research, prototyping, validation, and operations end to end. I shipped the Ghostie desktop agent, the AI Markdown tool Mdrone, automation agent kits for finance, and documented the information architecture, prompt templates, and rapid deployment methods behind custom agents. I regularly attack hackathons and online challenges, compressing product positioning, engineering, and growth experiments into 48-hour validation loops.

Six years in the Tianjin University School of Architecture plus hands-on work at the Zhejiang Provincial Institute of Architectural Design taught me to run system simulations and narrative design at the urban scale while mastering parametric modeling, material research, and user flow analysis. I drove multiple landmark projects from competition concept through construction drawings and began coding during this period, using Rhinoceros, Grasshopper, and Python for generative design, which later became the muscle behind my AI and product work.