When people ask me what I do, my response depends on to whom I'm speaking. When speaking to others who work in tech I'll usually say I'm a 'Data Scientist'; to most other people I'll call myself an 'AI Engineer'. When I need to fill in an online form, the most relevant dropdown option is usually 'Software Engineer'. The real answer is simply 'Engineer'. This actually answers a question much more fundamental than the one that's been asked because being an Engineer is not simply what I 'do', it's who I am.
Since January 2021, I've been working for CMR Surgical: a Cambridge-based surgical robotics company whose mission is 'to make the benefits of minimal access surgery available to everyone who needs it' (for what this means, see a very quick intro to robotic surgery).
In my first year, I worked as a Data Scientist in data analytics. This involved answering engineering and business questions with data: querying and consolidating large data sources, running statistical analyses on the data in order to draw out insight, publishing results into effective graphs, diagrams and summary. Findings were published either as technical reports on the process and its findings or as web dashboards for interactive data discovery. Another major part of this process was managing data pipelines for making our analyses quicker and repeatable. To do this, some of the tools I used were:
Following this I switched focus to work in Data Science for AI in an applied research capacity. This generally means researching, building, evaluating and testing datasets, machine learning models and MLOps deployments to a proof-of-concept level (and sometimes beyond). Robotics, as with many fields, is teeming with data therefore there is significant value that machine learning can add. I've worked on projects covering NLP, Computer Vision and Tabular Data as well as both Local and Cloud model deployments. Tools used for this not mentioned above include:
I've always been interested in robotics. This manifested itself early in the form of Lego Mindstorms at Primary School, Vex Robotics in Secondary School (UK School Champions 2015), and in my Master's year of my undergraduate degree at the University of Oxford. My research was facilitated by the Soft Robotics Laboratory at the Oxford Robotics Institute. I used an industry-standard robotic arm (Franka Emika Panda) fitted with a robotic skin to identify geometrically similar but materially different objects. More information about this can be found here.
Recently at CMR, I have been involved in improving DIB (Diverity, Inclusion and Belonging) through employee led groups as well as helping to found the CMR AI Club, which aims to promote knowledge, understanding and interest in AI and related subjects.
Outside of work, I tutor mathematics, physics, data science and programming from KS3 to Master's Degree level. Recently I have tutored two Master's students: one in quantitative finance with data science at UCL and one in data analytics at IDC Herzliya. Current students include an undergraduate in mathematics at the University of Chicago.
All content/graphics on this site were designed and built by me. All views expressed are solely my own.
Here's my understanding: minimal access/laparoscopic/keyhole surgery (all synonyms) is much better for everyone. It's better for patients because recovery times are quicker and less painful, it's better for hospitals because faster recovery means more available beds and it's better for surgeons because patients in less pain means happier patients. Except, it's not always better for surgeons because traditional keyhole surgery is very difficult (and very uncomfortable). This is where robotic surgery is superior, offering faster training, easier surgery, more comfort for the practitioner and therefore fewer early-retiring surgeons.