Designed and implemented a data pipeline to preprocess, geocluster, and predict rental revenues for properties on the market.Continuous evaluation and improvement of the pipeline above via CI checks, feature analyses, documentation, and refactoring.Conducted various analyses of the property market and created dashboards for investor presentations.Improved the asynchronous property listing scraper to adapt to new website designs.Lead the design of a pipeline to assess conditions of areas across the UK to find the best spots to buy properties from.

Refactoring and improving the software components of Pegasus that deal with user and organisation management, billing, and user authentication services. Pegasus is the main product of Visulytix which can analyse retinal imagery to provide decision support in diagnosis of sight threatening conditions. Significantly increasing code coverage via unit tests.Writing BASH scripts to automate tasks related to running unit and integration tests, and managing Docker containers.Heavily used libraries, software, and tools: Django, REST Framework, unittest, Docker, Amazon EC2, BASH scripting

Carried on responsibilities from previous Machine Learning Specialist role.Conducted exploratory data analyses for clients from different industries such as food manufacturing, and oil and gas.Developed end-to-end condition monitoring solutions for industrial machinery that involve data acquisition, machine learning, and dashboarding.Provided bespoke solutions for analysing and visualising data from GCloud.Heavily used libraries include NumPy, Pandas, Matplotlib, ZeroMQ, Dash, google-cloud, NI-DAQmx.

Delivered bespoke data acquisition and signal processing solutions in the manufacturing industry using Python.Developed condition monitoring and anomaly detection algorithms for industrial machinery and domestic appliances using National Instruments sensors and Verv, a smart home assistant developed by Austin Consultants’ sister company.Designed a Flask webapp dashboard to display condition monitoring results via interfacing with Amazon S3 in real time.Conducted business development endeavours.Heavily used libraries include Numpy, Matplotlib, scikit-learn, Flask, and ZeroMQ.