Aire Analytics

Background

Kevin’s background is in high energy physics where he measured differences between matter and antimatter, and nanosecond particle lifetimes. The large volume of data collected, and the lifetime-dependent efficiencies caused by the detector apparatus, resulted in a very challenging analysis. Each of the 200 million events had its own unique efficiency that had to be included in a maximum likelihood estimation.

Kevin went on to do natural language processing (NLP) of health triage data using recurrent neural networks. Vector embeddings were used, which are multidimensional representations of words constructed using unsupervised learning.

During the COVID-19 pandemic, Kevin was a key worker for the NHS, building the data warehouse and analysis platform used to land, load, analyse and disseminate test results and vaccination records. Kevin was also a data engineer tasked with creating a single asset of linked healthcare data, to be used by government and researchers to answer questions about COVID-19. He also worked on an NLP project to map textual medicine names to integer code (from the standard dictionary of medicines) to improve data quality and clinical safety.

Role

Data scientist, data engineer and software engineer. Particle physics PhD. Former research scientist on the LHCb experiment at the European centre for nuclear research (CERN).

Techologies

AWS
Python
Spark
Terraform
Docker

Methods

GBDTs
Maximum likelihood estimation
Spark
Terraform
Docker

Background

Zahid’s background is in data science applied into software development, and mathematics and statistics gained from a Bachelor of Science dual honours degree in astrophysics and chemistry.

Before joining Aire Logic, Zahid gained experience in analysing SQL databases and applying the findings to create bespoke reports using Microsoft Excel and PowerBI. He also used machine learning and artificial intelligence to fill-in gaps in client databases.

Zahid used geographical data to develop an algorithm that highlighted smart pallets that were not in the right locations. And developed a web application hosted on Microsoft Azure to provide reporting. The front end was written in Angular Node.Js that connected to SQL and NoSQL databases through a Python middleware which hosted the algorithm. The web application featured a mapping tool that the client used to recover lost pallets. As a result, in the first few months, Zahid’s client recovered around ~$10,000 worth of pallets per week.

Zahid has been involved in the migration of databases from SQL to the cloud in Microsoft Azure and restructuring data to be stored in NoSQL databases. He has also set up frameworks for clients to enable Big Data processing and storage.

At Aire Logic, Zahid works on the Decommissioning of National Health Applications and Infrastructure Services (NHAIS) programme. Zahid reports on the usage of the OpenExeter application which helps to identify and plan which areas of the NHAIS programme can be safely decommissioned. The reports have been created by cleansing data and presenting the findings using data visualisations created in Python and Excel pivot tables and charts.

Role

Data analyst, software developer

Techologies

Python, Scikit-Learn
R
SQL, NoSQL
HTML, JavaScript, Angular, Node.Js, TypeScript
Microsoft Azure, Excel, PowerBI

Methods

ClassMathematics and statistics
Data visualisation using Python and R
Data cleansing and manipulation
Supervised and unsupervised machine learning, neural network
Deduplication
Agile methodologies and DevOps

Aire Analytics

Source actionable insights and produce meaningful research from your data

Our services

Leverage your data.

Automate and drive value.

Predict risk and prognosis.

Manage your data.

Drive business intelligence.

Support data-driven decisions.

Meet some of our talented team

Connect with us

Stay connected