Technical capabilities

No percentage bars. Evidence instead.

"Python 95%" tells you nothing. Every capability below links to the work where it is actually used, so you can check the claim rather than take it. Only technologies that appear in work shown on this site are listed.

Explore

Pick a capability.

The matrix

Capability areas.

AreaTechnologies and methods
Software engineering

What production work taught me.

Most bioinformatics graduates have written analysis code. Fewer have had to keep a system running for other people. These are the specific things that came from doing the second.

01

Trust boundaries

ForgeLab grades code submitted by the person being graded. Anything the browser claims has to be re-established on the server, under a permission model that assumes the submitted code is hostile.

Node permission model · process isolation · bounded runtime

02

Shared architecture

Gradify runs 30+ sites from one kit. The rule that makes that survivable is that sites consume the kit and never fork it, so a fix lands everywhere at once.

ES modules · shared design system · no build step

03

Data that outlives the code

Version-controlled migrations, row-level security and a schema designed to be queried by someone who did not write it.

PostgreSQL · migrations · row-level security

Where I am still building. I would not yet claim production experience with workflow engines at scale, deep learning in biology, or cloud infrastructure beyond deployment. Those are listed in my proposals as directions, not in this matrix as capabilities. Overstating the range would be easy and would fail at the first technical interview.