Software Engineer
Stockport, England, United Kingdom
Working closely with the CTO and core ML engineers on a variety of greenfield software solutions in collaboration with clients, I was involved in developing (chronologically):- The CarbonFixers building design carbon estimation SaaS web-app, developed as a service-oriented PERN-GQL distributed stack hosted in the Azure cloud ecosystem.- Within CarbonFixers, a data-driven GraphQL API for the aggregation and calculation of construction carbon emissions from EPDs written in Python using the FastAPI and Strawberry-GraphQL frameworks.- A recommendation engine (p2p, c2p, and c2c) built as an Express.js API using the Gremlin query language to interrogate a knowledge graph built through a C# entity resolution ETL pipeline.- An AIaaS platform offering on-demand multilingual sentiment analysis built as a k8s cluster of request-response Python micro-services and a React-Apollo web-app leveraging HuggingFace transformer models served by OpenVino.- An AI-powered ETL pipeline, written as a dynamic Python CLI with Typer, for performing large-scale sentiment analysis on tweets retrieved from Twitter's REST API using BERT models.- A demand forecasting tool powered by a Temporal Fusion Transformer (TFT) model built using pola.rs for data pre-processing and PyTorch-Forecasting, PyTorch-Lightning, & Ray-Tuner for model definition, training, and tuning respectively.- An AIaaS image augmentation API as an event-driven Python/Go microservices cluster, with FastAPI as the API handler, deployed on Kubernetes leveraging HuggingFace CNN models served by OpenVino.