LOCAL INTERFACEClassroom knowledge system
I turn AI, research, and product ideas into systems you can verify.
I move from problem definition to implementation, testing and evidence. This portfolio shows not only the outcome, but also what I owned, what has been verified and what remains limited or conceptual.
VERIFIED LOCAL INTERFACE
Classroom Knowledge Reconstruction System
The project status and detail link remain available if the media fails.
Classroom Knowledge Reconstruction System
49 tests passed · Local frontend and backend verified · Sanitized interface evidence
View evidence and limits
SELECTED WORK / 01
Six projects. Six evidence states.
Software screens, tests, terminal output, structural photos and clearly labeled concepts form the portfolio. Work without public proof is not presented as complete.
TESTED CLI + DASHBOARDRisk reporting workflow
Team portalResearch areasLiterature learningAssessmentProjectsRecruitment
Public build evidence is still being prepared; no invented screenshot is used as a substitute.
CONCEPT
- Health records
- Lifestyle support
- Moxibustion scenario
- Multi-device concept
This is a product concept, not a validated medical product.
STRUCTURAL PROTOTYPECareRing
- 01Requirement framing
- 02Agent plan
- 03SolidWorks automation
- 04Patent drawing processing
Real task records, scripts, models and image-processing outputs form the case evidence.
VERIFIED LOCAL INTERFACEClassroom Knowledge Reconstruction SystemEvidence media unavailable
The frontend and backend have been verified locally with 49 tests passing. Public screens are privacy-reviewed and exclude student identities, API keys and source classroom data.
Open full case study ↗TESTED LOCAL WORKFLOWMarket analysis and risk reportingEvidence media unavailable
36 offline tests passed. The interface and reports demonstrate the workflow only. They are not investment advice, a profit claim or proof of autonomous order execution.
Open full case study ↗
Organizing research-team information, learning and project evidence into one entry point
Team portalResearch areasOutputsLiterature learningAssessmentRecruitment
The frontend structure has been reviewed. Public build evidence is still in preparation, so no invented interface is used.
Open full case study ↗
04 / PRODUCT CONCEPT
Product planning and user flow for health-management scenarios
- Health records
- Lifestyle support
- Moxibustion scenario
- Risk notices
- Multi-device concept
This is a product and competition concept, not a completed application or validated medical product.
Open full case study ↗STRUCTURAL PROTOTYPECareRingEvidence media unavailable
The photos and models show form, position and assembly intent. They do not validate electronics, sensing, atomization, safety or health effects.
Open full case study ↗
06 / WORKFLOW CASE COLLECTION
Making agent collaboration plan-driven, inspectable and reviewable
- 01Requirements and boundaries
- 02Stepwise execution plan
- 03SolidWorks automation
- 04Patent drawing and delivery review
Evidence comes from real task records, scripts, models and image-processing outputs. This is not one packaged software product.
Open full case study ↗Make the problem clear. Then make the result credible.
AI can accelerate implementation, but it does not replace requirement judgment, factual review, privacy boundaries or final acceptance.
- 01Define the real problem
Clarify the user, context, constraints and definition of done before selecting technology.
- 02Build a working path
Break the goal into pages, scripts, interfaces, models or structural prototypes with a fallback.
- 03Verify the output
Use tests, runtime screens, reports, models or structural photos to prove what actually exists.
- 04Document limits and next steps
State what is private, unvalidated or conceptual instead of hiding it behind vague language.
A rehabilitation-science background crossing into AI, web and engineering prototypes.
I am Liu Wenlong, also known as David Liu, a rehabilitation-therapy undergraduate and independent project builder. I participate in medical research and experimental coordination while building AI applications, research websites, automation scripts, market-analysis tools and smart-hardware structural prototypes.
My strength is not limited to calling one model. I turn ambiguous needs into executable paths while keeping facts, risk and delivery boundaries clear across disciplines.
Read the full public resume ↗- AI applications
- Model APIs, prompting, RAG foundations, feature flows and local prototypes
- Web delivery
- HTML, CSS, JavaScript, React, Next.js and static deployment
- Python automation
- CLI tools, data processing, tests, API calls and engineering scripts
- Product and engineering
- Requirements, interaction planning, CAD modeling, patent drawings and acceptance documentation
SELECTED RESPONSIBILITY
Deputy lead of a research experiment group · Competition project department lead · Requirements, implementation and test owner across multiple independent AI and web projects
CONTACT / 04
Need someone who can turn a complex problem into working evidence?
I am exploring opportunities in AI applications, web development, digital projects, product operations and cross-domain innovation. Review the project evidence, open the public resume or contact me directly.