Tomasz Kotarba

IT consultancy services on a contract basis: software engineering, web development, systems and network administration, databases and open source solutions. Please see below.

* 2018 to present: Morgan Stanley (London, UK)
Supplying software and systems engineering services for a tier one global investment bank. Four contract extension offers so far.

keywords: Python (incl. 2 to 3), GNU/Linux, DevOps, HTTP, streaming, message brokers, JSON, agile, greenfield, legacy, Scrum, Kanban


* eHealth Systems Africa (Nigeria / USA / Germany / UK - 2014 to 2015)
Working on multiple projects (most for UNICEF): three Django projects; one Android application; numerical computations and translation between Python, C# and JavaScript; designed a domain specific language for specifying complex questionnaires.

keywords: web development, Django, Python, TDD / test-driven development, AJAX, ECMAScript / JavaScript, JSON, PostgreSQL, databases, Kivy, mobile application development, GNU/Linux, systems administration, GIS (geographic information system)


* UCL (London, UK - 2012)
While providing UCL with consultancy service on software engineering, system architecture and web technologies I completed three independent projects which resulted in several recommendations, architectural changes to the existing infrastructure and 9 software products. The projects involved talking to clients, performing system analyses and two feasibility studies, producing recommendations and suggesting architectural changes, designing and implementing computer software tested on 10 different platforms and modifying legacy software. For my Client's opinion about the solutions I provided them with please see their recommendations on my LinkedIn profile.

keywords: UML, OO / object-oriented, component-based, Zope, Silva CMS, Plone, Python, AJAX, ECMAScript / JavaScript, XML, XSLT, HTML, CSS, databases, GNU/Linux, network security & systems administration

COMPUTATIONAL BIOLOGIST
This was my second project at the Center for Pathogen Evolution. Unlike the previous one, the main part of which consisted in software engineering, this project was heavily research-based and involved my applying techniques of bioinformatics and artificial intelligence to virology - more specifically, I proposed a solution which supplements a computational method known as antigenic cartography with pattern recognition capabilities of narrow AI to estimate the antigenic effect of genetic mutations in the seasonal influenza A (H3N2) virus. I created a method (and developed a working piece of software implementing that method used to run my computational experiments) which was able to position virus strains on antigenic maps with high accuracy (the coefficient of determination equal to 0.90-0.99 depending on the dataset) based solely on the primary structure of the antigenically important protein haemagglutinin acquired using the cost-effective and high-throughput molecular sequencing. Applying this method allows researchers to update existing antigenic maps with virus strains for which the labour-intensive and time-consuming process of antigenic characterisation using serological assays has not been performed.

keywords: artificial intelligence, machine learning, data analysis, statistics, computational biology, virology, bioinformatics, Python, NumPy, SciPy, OO / object-oriented, GNU/Linux, matplotlib, SVM (support vector machines), ANN (artificial neural networks), hyperparameter optimisation, protein sequence databases, grid search


RESEARCH ASSOCIATE / SOFTWARE ENGINEER
To give back to the society, I decided to temporarily become a terribly underpaid science monk and apply my software engineering and computational skills to help good people in academia save lives and fight Influenza (a disease which sadly kills around half a million people every year).
During my first project (funded by the US CDC / CEIRS) at the Center of Pathogen Evolution, I engineered a system, the main purpose of which was to aid experts in understanding, assessing, and mitigating, the risk of avian A/H5N1 influenza viruses emerging in a mammalian host. The system achieves this by monitoring publicly accessible genetic data sources for newly emerging virus strains, performing and analysing multiple sequence alignments on thousands of nucleotide sequences, reconstructing phylogenetic trees, annotating the trees with mutations of interest (including those specified in the US CDC H5N1 Genetic Changes Inventory) and clades (according to the unified nomenclature for the highly pathogenic H5N1 avian influenza viruses by WHO / OIE / FAO), and producing interactive reports by visualising results using trees, molecular structures, time-series, PostScript, HTML, graphics and JavaScript. My work involved all phases of developing this software: interacting with domain experts (who worked for/with WHO Global Influenza Surveillance Network), analysing the problem domain and specifying system requirements, designing, implementing, testing, deploying and maintaining the system and helping my successors to take over after I completed my work and moved to my next project. Some parts of the system were also applied in research projects concerning the Dengue and H7N9 viruses.

keywords: Python, bioinformatics, NumPy, SciPy, bash, GNU/Linux, OS X, systems administration, PostScript, HTML, CSS, ECMAScript / JavaScript, Lisp, MAFFT, PhyML, PAUP, GARLI, virology, OO / object-oriented, databases

Computational Biology Approach to Cancer Virotherapy - an independent research project (conducted for Oncos Therapeutics & CGTG (Cancer Gene Therapy Group / University of Helsinki)) applying complex systems theory, machine learning and multivariate data analysis to the problem domain and conducted as part of my MRes degree. Subject to non-disclosure agreement; no further details available.

keywords: Python, bioinformatics, NumPy, SciPy, virology, gene therapy, OO / object-oriented, computational biology, databases, data imputation, machine learning

Planned and conducted the following research projects (each of them produced working software as a side effect):
■ Object-Oriented Executable Biology Framework for Prostate Cancer Modelling
■ In silico method for analysing patterns in spatial arrangement between TIM barrel domains and other domains inserted into the same protein chain

keywords: Python, bioinformatics, NumPy, SciPy, matplotlib, OO / object-oriented, sage / sagemath, computational biology, protein databases

Following Land Securities Group's (one of the FTSE100 corporations and the then world’s third largest real estate investment trust) acquisition of SMIF in February 2007, and after the successful integration with the new company's ICT infrastructure, my main responsibility was to maintain and further develop the specialized PPP/PFI management information system I single-handedly created for SMIF (main focus: business intelligence). Apart from the virtualisation platform (VMware) I used as a foundation for the load-balancing cluster with fail-over capabilities to host the system, all technologies employed by me to build the system were Open Source.

keywords: object-oriented software engineering, OOA, OOD, OOP, component-based architecture, Debian GNU/Linux, ZODB (Zope Object Database), ZEO (Zope Enterprise Objects), Zope (Z Object Publishing Environment), application servers, Apache, web servers, CMF (Content Management Framework), Plone, CMS (content management systems), finance, liquidity, infrastructure investment, PPP (Public-Private Partnership), PFI (Private Finance Initiative), web development, Python

I was in charge of ICT (Information and Communications Technology) at SMIF (Secondary Market Infrastructure Fund) which was at that time the biggest secondary private finance initiative investor in Europe and one of the largest European infrastructure investment and management groups. It was a leading provider of liquidity to infrastructure project investors and developers in the UK and European PFI/PPP market. Some of my responsibilities: providing and maintaining secure and reliable network connectivity between all company offices; developing and maintaining the company information system; managing IT staff and contractors; information security etc.

keywords: Debian GNU/Linux, Python, ZODB, ZEO, Zope, application servers, middleware, SQL, NoSQL, Apache, web servers, CMF, Plone, CMS, finance, liquidity, infrastructure investment, PPP / PFI, network protocols, TCP/IP, VPN, IPsec, NTP, FTP, SSH, HTTP, SSL, TLS, firewalls, routing, storage, information security, computer networking, computer hardware, network administration, systems administration, samba / CIFS, LDAP, Microsoft Windows, active directory, Microsoft Exchange, IT management, web development, BI (business intelligence)