Alex McRoberts

Grew and led a team from 4 working on Employee Advocacy to a portfolio of 16 developers across two teams. Projects I led teams through included migrating from EC2 to a fully hosted Kubernetes solution, migrating to a hosted MongoDB solution, migrating from Python 2 to Python 3, a full rewrite of the Web UI, and a migration from a hybrid mobile app to the native Hootsuite mobile apps.

Working in a lead engineer capacity across a multidisciplinary team of hardware, software and mechanical engineers, leading Research & Development into automotive safety with sensors, machine learning and cloud connectivity.

As the Headworn Apps Lead Engineer, I lead a team that worked closely with customers and partners to deploy Intel Head Worn solutions in their enterprise environments. This involved understanding requirements, problem solving, bug fixing, and developing solutions for the head worn devices suited to the individual needs of the enterprise, working closely in conjunction with Business Development teams. I often travelled to sites to help understand how we could best deploy a solution for the customer. Daily tasks involved working on apps with Wi-Fi, Bluetooth, and sensor data from Internet of Things (IoT) devices via Bluetooth Low Energy (BLE). Other apps include custom User Interface and User Experience, and interacting with REST APIs.

As the lead of the web team, our focus was on rebuilding and growing the web backend that powered data synchronization and software updates to the Recon Snow2 and Jet wearable Heads Up Displays. I lead the refactor of the PHP backend using a MySQL database, driven on the front end powered by JavaScript libraries including Backbone.js, Underscore.js, RequireJS and jQuery. I architected APIs which controlled software updates, data uploading and processing, and an OAuth 2.0 based login process for desktop software which used those APIs. I also lead the refactoring of a Java Spring backend tied to a Postgres Database with PostGIS, which involved processing thousands of records of the recorded GPS data from the HUD, to return meta data such as Ski Trail names, min, max, average stats about a user's day, and to summarize data to make it presentable on the front end. As part of the Java refactor, I built a RabbitMQ / AWS S3 based queueing system to allow processing tasks to become asynchronous. This allowed APIs consumed by Mobile Apps to close radio connections faster, allowing significant battery savings on Mobile Devices. I introduced a novel battery saving technique by building an asynchronous Push Notification System that allowed Mobile Devices to rely on Push Notifications to power the API driven data lifecycle within the mobile apps - user coming online, going offline and uploading data.

Our focus was on rebuilding and growing the web backend that powered data synchronization and software updates to the Recon Snow2 and Jet wearable Heads Up Displays. Working with the team I helped refactor the PHP backend using a MySQL database, driven on the front end powered by JavaScript libraries including Backbone.js, Underscore.js, RequireJS and jQuery. I architected APIs which controlled software updates, data uploading and processing, and an OAuth 2.0 based login process for desktop software which used those APIs. I also refactored a Java Spring backend tied to a Postgres Database with PostGIS, which involved processing thousands of records of the recorded GPS data from the HUD, to return meta data such as Ski Trail names, min, max, average stats about a user's day, and to summarize data to make it presentable on the front end. As part of the Java refactor, I helped build a RabbitMQ / AWS S3 based queueing system to allow processing tasks to become asynchronous. This allowed APIs consumed by Mobile Apps to close radio connections faster, allowing significant battery savings on Mobile Devices. I introduced other battery saving techniques by building an asynchronous Push Notification System that allowed Mobile Devices to rely on Push Notifications to the API lifecycle within the mobile apps.

The Brain Computer Interface program focused on developing BCIs with Rapid Automated Interfaces for Nonexperts (BRAIN) which aimed to develop Brain-Computer Interfaces (BCIs) into practical assistive and ICT tools to enhance inclusion for a range of different disabled users. Under the guidance of Dr. Paul McCullagh and Dr. Gaye Lightbody, I developed software including a novel BCI digital drawing tool in the style of an Etch-a-Sketch. As a user interacted with an BCI SSVEP system using 4 lights that flickered at different frequencies the system could detect which way the line should be drawn, enabling a user who had no motor functionality in their arms to draw on their computer. I worked remotely with colleagues based in Spain and the Czech Republic. We developed a Zigbee based Smart Home system where we could reliably control a Smart Home using a BCI system. This would allow a user to turn lights on and off, open and close doors, and turn electronic devices - such as a home radio off and on. Written in Java - using the OSGi framework with UPnP technologies I was able to demonstrate control of a Digital Media Player with functionality such as skipping songs, selecting and playing videos and changing volume from a client's machine over an Xbox running on a local network. Other tasks including running BCI testing sessions with clients from the Cedar Foundation alongside Dr. Melanie Ware and Dr. Suzanne Martin. This involved working with a trained BCI Technician who would adminster the BCI electrodes to a client, before we ran our tests which involved evoking an SSVEP from a client - getting them to focus on a light source for a short period of time. This work lead to two published papers - including one printed in the HCI International Journal in 2011.

I worked on several projects under several several professors. The Virtual Mentor project under Dr. Juan Carlos and Dr. Paul McCullagh. The purpose of the project was to demonstrate how a Virtual Mentor chat bot could provide support to new students in a campus environment. By merging an Artificial Intelligence based chat bot with a speaking Avatar the hope was that the animated character on screen could provide more relevant help to a student, compared to trawling through a website in a traditional manner of searching and clicking. By building a database of question / answer pairs, we could demonstrate the ability of a system to respond in a timely manner using a speech engine, and by displaying text on screen. A brief demonstration of the system showed that such a system could provide feedback to a student in a more friendly manner, allowing them to feel more informed, and reassured about the information the received. Scoping the Potential for Anytime-Anywhere Support Through Virtual Mentors - Juan Carlos Augusto, Paul McCullagh, and Alex McRoberts. 11th Higher Education Academy - ICS Annual Conference, Durham University, 24th August - 26th August 2010. Read More: http://journals.heacademy.ac.uk/doi/abs/10.11120/ital.2010.09020010

Working under the guidance of the Meridio Chief Architect in the R&D department of Meridio - an eDRM product, I was tasked with prototyping software for proof of concept demonstrations that lead to the integration of a new Search Engine into the Meridio eDRM system. During the final 6 months of my internship my role pivoted from R&D to Sales Engineering. Working with a mixed team of Developers and Sales Managers to produce demonstrable prototypes of features which where specifically tailored to clients needs. This work allowed me to develop great communication skills, and the ability to meet tight deadlines in an environment of constantly changing requirements.