Laboratory science × software systems
Biotechnologist during the day,software developer at night.
I’ve been programming since I was fourteen. Today, I combine that background with biotechnology and laboratory research to build scientific software, supervised AI tools, and practical systems.

Evidence behind the work
Not just claims. Actual context.
The projects are grounded in laboratory training, published research, and a long-running interest in building software.
Biotechnology degree
B.Sc. Biotechnology, Hochschule Rhein-Waal, completed in 2025.
Wet-lab experience
Cell culture, Western blot, flow cytometry, viability analysis and IC50 workflows.
Published contribution
Co-author on a 2025 pleural mesothelioma study in Cells.
Programming since fourteen
C, C++, C#, Lua, Visual Basic, embedded C, Python, R and Java for Android.
CURRENT RESEARCH
Help me understand what slows laboratory work down.
I’m collecting feedback on everyday analysis and data-handling pain points in laboratory research. Your input will help shape the scientific tools I build next.
Selected projects
One laboratory tool. One personal assistant.
Each project gets a proper case study, ongoing changelog, and related articles instead of being reduced to a tiny portfolio card.
Dose-response comparison
PF155 · PF760 · PF998

MesoLab Analytics
An offline analysis environment designed around actual drug-response workflows rather than generic spreadsheet logic.
- Guided plate layout, quality control, normalization and 4PL fitting.
- Designed for researchers who should not need to become data engineers.
- Grounded in mesothelioma research and real laboratory constraints.
Approve code change?
Suggested after reading current workspace context
AEDA
A local-first assistant built around explicit permission, visible tool activity, and supervised changes.
- Permission broker, dry runs, unified diffs and rollback.
- Hardware-aware local model selection with cloud fallback.
- Wraps mature tools instead of pretending to replace everything.
Draft → review → publish
Hermes writes. You edit. Hermes publishes after approval.
The blog workflow is designed around explicit review, not automatic posting without oversight.
Projects that explain themselves
Case studies hold the polished overview. Blog posts document progress, failures, design decisions and changelogs.
About preview
Full story on the About page: Yavar Dezfouli.
I started programming at fourteen and kept moving between languages, platforms and small experiments long before studying biotechnology.
The full About page connects that history to laboratory work, publication experience, the origin of MesoLab, and why AEDA focuses on helping people use existing tools safely.
From the blog
Recent articles.
These four initial articles are written but intentionally marked as drafts until you approve their wording.
Introducing MesoLab Analytics
Why real wet-lab workflows need something between spreadsheets and heavyweight analysis suites.
Read articleWhy AEDA assists instead of replacing mature tools
The case for wrappers, supervision, and choosing the right existing tool.
Read articleBuilding a supervised permission system
Dry runs, scoped approval, visible tool activity, unified diffs and rollback.
Read articleBuilding the yavard.de self-hosted stack
Lessons from Docker, Caddy, mail, monitoring and service operations—without exposing private systems.
Read articleContact
Science, software, or something between them?
Reach out about biotechnology, laboratory tools, local-first AI, research software, or a useful collaboration.