Business systems and automation specialist with a project management background in AI product development. I translate operational needs into requirements, workflows and practical prototypes, connecting business goals with technical decisions. Experience spans enterprise process automation, CRM and KYC workflow design, AI applications, and cross-border collaboration in Chinese, English and Japanese.
Led Prometheus Lab AI, an AI-assisted laboratory workflow monitoring product, from requirement definition through technical integration, coordinating a six-person product team spanning AI, engineering, software/hardware integration, marketing and business development, alongside management and external partners.
Drove the product into the Final Round of the Qualcomm Innovate in Taiwan Challenge 2025, owning the proposal, progress control, technical documentation and competition deliverables; the product was later shown at CES 2026, Smart City Summit & Expo 2026 and Extreme Tech Challenge 2026 Japan.
Established the product direction for first-person laboratory workflow recognition by integrating smart glasses with AI modules, and coordinated international collaboration with multiple medical and research institutions across three countries.
Built n8n automations connecting Ragic, Outlook, Excel and government and other public APIs across three areas — general administration, several AI development projects, and marketing (AI product marketing and SEO/GEO keyword automation) — applying predefined decision rules to route and summarise information, saving an estimated 20 hours of manual work per week.
Designed the content architecture and AI generation workflow for a cancer-care SEO programme — prompt design, output format, medical review logic and data verification — supporting content production across more than ten websites.
Wrote and integrated proposals for SBIR, the National Science and Technology Council and AAMA, and served as the Chinese / English / Japanese interface for international meetings and technical presentations.
Supported 20+ company projects covering data consolidation, cross-team communication, SOP design and process standardisation.
Analysed and visualised tens of thousands of records on physician prescribing trends and multi-cancer treatment statistics, producing the reference data used in management reporting.
Represented the company at international exhibitions including SusHi Tech Tokyo, presenting products in Chinese, English and Japanese; completed 80+ translation, subtitling and presentation localisation items.
Tested internal CRM and reporting systems, reproduced and reported defects, and corrected HTML issues affecting report output.
Supported research reporting, study protocol execution and statistical analysis.
04
Career transition
11/2025 – 06/2026
Planned transition period focused on repositioning toward business systems, automation and technology-driven services.
03
Selected projects
05
AI Control Center — orchestration for AI coding agents
08/2026 – present
Designed and built a Windows desktop application that coordinates AI coding agents through defined roles with duty-bound permissions, verification steps and human approval, with an audit timeline for every project.
Set the verification discipline: 2,200+ automated tests under a zero-skip rule, and 860+ deliberate code breaks that check whether a test can fail; a sweep with this method found 30 false greens and cleared them to zero. Source: github.com/SeanSyue/ai-control-center (MIT). An earlier build was installed, launched and uninstalled on Windows; known limits are documented in the repository.
06
Enterprise KYC / CRM and case management system
07/2026 – 08/2026
Redesigned the end-to-end workflow from preliminary KYC and needs analysis through quotation, advanced KYC and contracting, specifying forms, data structures, document filing and role-based permissions.
Built a local case management MVP in SQLite covering case records, role permissions, case isolation, backup and verification, and completed dry-run testing for cross-case access, duplicate submission, date formats and data integrity.
07
AI production pipelines — image, music and game
2026
Turned subjective art direction into a written specification and an automated colour check, after measuring that a similarity score could not distinguish a correct result from a wrong one.
Produced a 76-track game soundtrack in two days, with every render verified automatically before review and an honest test showing the screening script did not predict human judgement.
Directed a Godot life-sim in which a locally run language model proposes character actions that the game's rules validate before they happen; selected the 4B model after a test on the target PC: 0.53 s median reply against 3.26 s for the 9B, which did not fit fully in GPU memory, with no clear quality advantage for the 9B across eight test scenarios.
04
Skills
Business analysis & process
Business Process Automation · Requirements Analysis · Workflow Design · CRM Systems · KYC / Client Onboarding · SOP Design · Cross-functional Coordination
Automation & integration
n8n · REST API · Webhooks · Google Apps Script · Ragic · Excel / Outlook / OneDrive
AI
AI agents · Claude Code · OpenAI Codex · MCP · Prompt engineering · Local LLM deployment and evaluation · ComfyUI