Biostatistics consulting
Study design, sample size and power, analysis plans, and statistical modelling — done to a standard you can defend to collaborators, reviewers, or decision-makers.
The real problem
Your data is too sensitive for careless automation.
Many teams want the speed of modern AI and automation, but their work depends on sensitive data, institutional context, and decisions that cannot be delegated blindly. The question is not simply whether to use AI. The question is how to design a boundary that keeps the data, the assumptions, and the final judgment under control.
A different approach
Machines can assist. Humans still decide.
I design workflows where machines help with structure, coding, triage, reporting, and documentation, while people remain responsible for the question, the context, the evidence standard, and the final decision.
Speed from machines. Judgment from humans.
The goal is not to remove people from analysis. It is to give them safer systems, clearer evidence, better defaults, and more time to think.
Services
Four ways to work with me.
Data pipelines & R tooling
Reusable data cleaning, validation, reporting, Shiny apps, R packages, and internal tools your team can run, inspect, and maintain.
AI-assisted workflow design
Practical, human-in-the-loop workflows for using AI safely in analytic work: scoped context, clear boundaries, review gates, and reproducible outputs.
Statistical review & second opinion
A focused review of an analysis, methodology, workflow, or vendor report: what holds, what is fragile, and what should be checked next.
What I build
Transparent deliverables your team can keep.
Everything I build is designed to be read, understood, validated, and maintained — not hidden behind a black box.
- Reusable data cleaning pipelines
- Validation rules and data quality checks
- R packages and internal utilities
- Quarto reports and automated reporting workflows
- Shiny and lightweight web tools
- Spreadsheet-to-pipeline transitions
- AI-assisted analysis workflows with human review
- Documentation and SOPs for recurring data work
My philosophy
Evidence first. Automation second.
A workflow is useful only if people can understand what it is doing, why it is doing it, where it might fail, and who remains accountable for the decision. AI can help with the work, but it should not hide the reasoning.
Humans frame the questionMachines can assist the work, but people define the decision, the context, and the success criteria.
Assumptions stay visibleModels and automation are useful only when their limits, uncertainty, and failure modes are clear.
Sensitive context is minimizedGood systems expose only what is needed and keep the most sensitive material behind the proper boundary.
About
Lennon Li
I am a biostatistician by training, with a background in statistics, computer science, public health, and applied data systems. I have spent more than 15 years building statistical analyses, surveillance models, R tools, and reproducible workflows for teams working with complex and sensitive data.
My work is guided by a simple belief: data should help people see more clearly, not replace their judgment. Models, dashboards, and AI systems are useful only when their assumptions, limits, and uncertainty remain visible.
Today, I focus on trustworthy machine-assisted decision making — using statistics, software, and AI-assisted workflows to help teams work faster while keeping humans responsible for context, meaning, and final decisions.
Process
How I work
- Understand the question — what decision will this analysis or tool support?
- Set the data boundary — what tools may see, what must stay local, and what can be safely summarized or simulated?
- Build the workflow — the code, model, report, or tool is developed as a reusable system, not a one-off file cleanup.
- Validate through gates — assumptions, outputs, and failure modes are checked with explicit review and validation steps.
- Review and hand over — the final deliverable includes readable code, plain documentation, and a workflow your team can keep.
The shift
From “clean my file” to “build my workflow”
Cleaning one spreadsheet may solve today’s problem. Building a workflow prevents the same problem from returning next week.
Contact
Need safer, more sustainable data work?
If your team depends on repeated reports, exported files, spreadsheets, manual cleanup, or fragile analysis workflows, I can help turn that process into something reusable and defensible.
Please do not send sensitive data by email. We can first discuss your workflow and decide on an appropriate secure process.