At my workplace, I majorly do all development in Rust. I am comfortable with Java and Python, of which I use the latter to pursue my interests in AI/ML algorithms. I am familiar with SQL, for which I did a couple of DataCamp certifications.
During my postdoc, I used the Python-based mathematical software system Sagemath and was amazed at its user-friendliness compared to Macaulay. I used Sagemath to work out proofs of concept for all the conjectures my advisor and I had. However sometimes it was not enough, we needed numerical solutions for systems of polynomial equations. We used Bertini for this purpose, which is a software package written in C, with simple syntax.
Like most academics, I have written my PhD dissertation using LaTeX for typesetting. I have and still use a LaTeX library, TikZ for drawing scientific diagrams in patent application drafts and so on. Recently I have cultivated the wonderful habit of quickly chalking up workflows for any project using UML diagram syntax from PlantUML and the easy-to-share interface Niolesk.
I have been working at numerous open-ended questions concerning our scaleable, end-point monitoring cybersecurity software. One of my recent projects concerns designing Word to Vector algorithms and testing the performance of existing locality-preserving word embeddings on the specific kind of data we need to compress.
I also have been looking at the machine learning algorithms currently employed in our product and obtaining a deep enough understanding before we can deploy the enhanced versions of these algorithms. My current research involves design and analysis of algorithms, practicalities of distributed systems, and the behavioural science of the software system underlying our product.
I am a big proponent of documentation of all work. To this end, I have been heavily documenting all my development work (Rust projects). I am currently working on patent applications for the core IP of my company, which has been a rich experience so far.
