Roderick Hodgson, Developer in London, United Kingdom
Roderick is available for hire
Hire Roderick

Roderick Hodgson

Verified Expert  in Engineering

Machine Learning Developer

London, United Kingdom

Toptal member since October 27, 2016

Expertise
VoIPPythonC++AWSWebRTCMachine LearningRecommendation SystemsWeb DevelopmentMySQLDjangoAgile DevelopmentCREST APINode-RED
Bio

Roderick is a full-stack developer with over a decade of applied research and product development experience focused on video streaming, computer vision, and natural language processing. Roderick has taken a leadership role in delivering projects ranging from a machine learning platform deployed at scale for the BBC to several innovative real-time audio/video products for which he designed and developed both the software and cloud architecture.

Portfolio

SQR Systems Ltd.
Amazon Web Services (AWS), MySQL, Stripe, SIP, Node.js, MQTT, React, C++11, C...
SQR Systems Ltd.
MySQL, Python, RTP, Streaming, Android NDK, Android, C++11, C, Video Codecs
BBC R&D
Amazon Web Services (AWS), CODE, Bundler, Embedded Systems, Xapian, Arduino...

Experience

  • Python - 12 years
  • Machine Learning - 10 years
  • Video Streaming - 9 years
  • C++ - 9 years
  • Amazon Web Services (AWS) - 5 years
  • Node.js - 5 years
  • VoIP - 4 years
  • React - 4 years

Availability

Part-time

Preferred Environment

Amazon Web Services (AWS), IntelliJ IDEA, PyCharm, Atom, Ubuntu, Git, MacOS

The most amazing...

...project I've worked on was a system that was able to automatically identify keywords in over 20,000 hours of audio recordings in a matter of days.

Work Experience

Head of Research and Development

2014 - 2016
SQR Systems Ltd.
  • Created new technology in C++ combining computer vision with video compression. Used the outcome of computer vision algorithms as input variables to the video encoding/decoding algorithms. Communicated feature-sets between sensors through a mesh organization.
  • Developed back-end microsystems for account management and payment systems in Python (Flask) and PHP Slim, integrating with Stripe.
  • Managed and delivered a number of projects to customers—from bidding process through to final delivery.
  • Chose and built technology to answer specific, immediate business requirements.
  • Established the technical innovation strategy and translated this into the product development roadmap with the head of engineering.
  • Determined the direction of the long-term mobile app product roadmap and worked with the head of engineering to translate it to sprint planning.
Technologies: Amazon Web Services (AWS), MySQL, Stripe, SIP, Node.js, MQTT, React, C++11, C, NDK, Android, VoIP, Flask, Python, Ansible, REST APIs, RESTful Development, Opus, H.264, Image Processing, Internet of Things (IoT)

Software Developer

2013 - 2014
SQR Systems Ltd.
  • Created the initial architecture of C++ video compression and encryption codebase upon joining the company as the second employee.
  • Maintained code documentation and code testing processes.
  • Developed novel algorithms combining video compression and data encryption in C++.
  • Brought encryption into the heart of the video compression pipeline.
  • Created the first Android app prototype. Used novel IP in the C++ codebase, through the Android NDK.
  • Wrote project bids for a number of research funding projects and commercial contracts.
Technologies: MySQL, Python, RTP, Streaming, Android NDK, Android, C++11, C, Video Codecs

Research Engineer

2009 - 2013
BBC R&D
  • Developed a visual way of exploring similarities between program segments in JavaScript by weighing keywords found in subtitles; clustering the results using non-negative matrix factorization in Python over results found with Xapian.
  • Implemented a RESTful API for portability to other projects.
  • Created an automated solution for linked data tagging of over 70,000 radio programs using Amazon Web Services and RabbitMQ. CMU Sphinx was used for speech-to-text, combined with text processing and disambiguation using the topology of DBpedia tags.
  • Developed a web front end and an API in Python for control and logging of the automated system.
  • Researched a new concept for immersive 3D environment exploration called "The Holohelmet."
  • Built a bundle-adjustment based solution for determining the user's position in a room using a small camera placed on a pair of headphones, and clusters of infra-red LEDs attached to the ceiling as markers.
  • Added an augmented reality projected over the environment using a pico-projector.
  • Co-authored a patent on a solution for determining program start time from an EPG-based recording, by correlating incidental signals in the DVB transport stream using a global alignment algorithm.
  • Implemented a set-top box VM, moving the algorithm offsite, as a centralized RESTful API.
Technologies: Amazon Web Services (AWS), CODE, Bundler, Embedded Systems, Xapian, Arduino, AMQP, Speech to Text, Image Processing, Text Processing, Machine Learning

Contract Developer (Freelance)

2008 - 2008
Robertson Technology
  • Scoped-out, designed, and developed a ticketing web app for use by a managed services company.
Technologies: JavaScript, PHP

Contract Developer (Freelance)

2007 - 2007
Owen Russell Group
  • Developed a data room intranet web application product for the BBC.
  • Built a change management intranet web application product for the BBC.
  • Performed a risk analysis and created a strategy for migrating other product codebases to the latest version of ASP.NET.
Technologies: Microsoft SQL Server, ASP.NET

Co-founder | Web Developer

2002 - 2005
FluideConcept
  • Founded a web development consultancy while in a high school in Switzerland.
  • Developed a number of websites for local businesses focusing on the public web presence and intranet portals.
  • Worked with a development stack primarily made up of PHP and MySQL.
Technologies: MySQL, PHP

Experience

Automated Semantic Tagging of Speech Audio

http://thenextweb.com/video/2012/09/17/tnw-ibc-the-bbc-describes-automatic-tagging-world-service-archive/
A project aimed at making the 60 years (20,000+ hours) of radio programs from the BBC World Archive searchable and browsable. This was done using speech detection, keyword extraction, and tag disambiguation. The algorithm was distributed over a large number of powerful AWS machines.

Free-Viewpoint Visualization

http://www.bbc.co.uk/rd/blog/2011-03-rd-and-blue-peter-ski-rossend
A project aimed at generating free-viewpoint video using 16 cameras. I developed part of the camera calibration system, which was used to determine the exact position and orientation of the cameras as well as the focal length and distortions.

Open Standards for Video Communications in Military Operations

A project with the UK Ministry of Defence, to trial interoperable solutions for video streaming in constrained environments. It runs with custom video encoding solutions on off-the-shelf single-board computers.

Ceerus Secure Voice and Video Calling Mobile App

The Ceerus mobile app was a project to bring a unique combination of compression and encryption, to provide more reliable secure calls. All while interoperating with existing technologies.

Flip Cam Mobile App

https://techcrunch.com/video/flip-cam/519307642/
Flip Cam was a TechCrunch Disrupt Hackathon iOS app that lets users instantly team-up with passers-by to create amazing multi-camera videos of interesting events. Uses Node.js and MongoDB, with video transcoding jobs performed by FFmpeg, queued through RabbitMQ.

Distillate

https://youtu.be/1KhA6_2Ov3w
Distillate was a TADHack 2016 hackathon project that allows users to record and track key topics from voice conversations over WebRTC. This allows users insight into what topics were discussed, when, and with whom. Uses the Telestax RestComm system, Node.js, and IBM Bluemix Machine Learning APIs.

Source Protector

https://www.youtube.com/watch?v=vM2VnTVnWnA
Source Protector was the outcome of the TADHack 2015 hackathon. It provides journalists with a dedicated phone number where their sources can reach them. Each of the sources' voices is distorted to preserve anonymity. It uses the Dialogic® PowerMedia™ XMS system and custom Python code for audio signal processing.

Education

2005 - 2009

Bachelor of Science Degree with Honors in Artificial Intelligence and Computer Science

University of Edinburgh - Edinburgh, UK

Skills

Libraries/APIs

WebRTC, REST APIs, React, PJSIP, AMQP, Stripe, Node.js, NDK, TensorFlow, OpenCV

Tools

Ansible, Android NDK, Node-RED, MQTT, Git, Atom, PyCharm, IntelliJ IDEA, Bundler, RabbitMQ

Languages

C++, Python, JavaScript, C, C++11, Java, PHP, SQL

Platforms

Amazon Web Services (AWS), Web, Arduino, MacOS, Ubuntu, Android

Frameworks

Unity, Flask, Django, ASP.NET, CODE, Android SDK, Apache Spark

Paradigms

RESTful Development, Web App Design, Model View Controller (MVC), Agile Software Development, REST

Industry Expertise

Web Design

Storage

MySQL, Microsoft SQL Server, Xapian, Elasticsearch, MongoDB, SlimPHP

Other

SIP, VoIP, Audio Streaming, Video Streaming, Audio Codecs, Video Codecs, H.264, Cloud, Recommendation Systems, IoT Protocols, Technical Project Management, Text Processing, Opus, Image Processing, Machine Learning, Internet of Things (IoT), Streaming, RTP, Speech to Text, Torch, Embedded Systems

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring