Keenan Ikking

At BioInformatiCo, I built a multi-stage Java pipeline that turns raw PDF geometry into structured logical sections. I also redesigned file uploads so new types can be added through configuration rather than shared-code changes.

I built a testing environment to distinguish retrieval misses from LLM misreads, helping raise usable trial protocols from 7 to 19. I introduced reference outputs and used precision, sensitivity, and F1 scores to validate pipeline results.

I designed an R-based DataFlow framework for clinical-trial reporting and programmed Safety Summary and Safety History reports through UAT cycles. For a difficult clinical report, I used test-driven development and a single test row mutated across 21 assertions to support a fast build and absorb QC fixes without regressions.

At the University of the Witwatersrand's Gentle Lab, I built and published nano-pipe, an open-source long-read RNA analysis pipeline, and developed benchmarking and evaluation pipelines for isoform-discovery tools. I have an MSc in Applied Bioinformatics with distinction, and I'm looking for a backend or data engineering role tackling hard problems in pipelines, design, and performance.