What does it take to work in machine learning? Sneha Sen, 27, is a senior machine learning engineer at London-based financial technology (fin-tech) company Cleo.
By her own admission, Sneha’s path to her chosen career was unconventional. Many of her colleagues who work in technology have degrees in maths or computer science. But at school, Sneha’s passion was languages.
She took three languages at GCSEThe national exams taken by 15 and 16-year-olds in parts of the UK. , followed by A-levels in French, German and maths. She was fascinated by how languages work. “I just got really into grammar,” she says.
Completing an extended projectAn extended project qualification (EPQ) is an optional, self-directed research project that students in the UK can take alongside A-levels. It is worth 50% of an A-level. on how children learn languages, Sneha discovered linguistics, the scientific study of language. She knew she wanted to study at Cambridge — and so she set about doing everything she could to learn more about the course.
After emailing a Cambridge linguistics professor, they introduced her to students studying linguistics, whilst a UCLUniversity College London, one of the top universities in the UK. professor offered her a week of work experience.
At this point, Sneha did not know what she wanted to do with a linguistics degree. She forged ahead anyway. “The jobs you can get might not even exist right now,” she explains. “So it’s best to just be interested in your subject and trust that things will work out.”
In her second year at university, she took a module in computational linguistics, exploring algorithmsAny set of rules followed by a computer. In the context of social media, “the algorithm” refers to the intelligent AI that learns the interests of the user and presents them with posts that it thinks will interest them. and how computers learn language. That summer, she completed an internship at the Computer Laboratory in Cambridge. It was her first introduction to PythonA programming language that is often used in web applications, software development, data science and machine learning. coding, and something clicked.
When she began applying for graduate schemes the following year, she knew that she wanted to use her linguistics skills in a practical way. She was taken on as a tech trainee data scientistA new type of data expert. Data scientists work with businesses to understand data and help to convert it into action that will improve a business or organisation’s performance. at Education FirstOften known as EF, an international education company specialising in language learning. , using natural language processingA subfield of computer science and AI. It uses machine learning to allow computers to understand and communicate with human language. to explore the errors language learners make.
Then disaster struck. Six months after she started her first job, the pandemic hit. Sneha’s whole team was made redundantNot needed any more. Someone who is made redundant loses their job because their work is not needed. .
She used her time to take free courses online in algebra, statistics and machine learningA field of artificial intelligence that aims to use data to teach machines to “learn” for themselves without the need for specific programming.. Soon she had a new data science job at an insurance company, before taking on roles at a language learning app and then a small company creating teaching resources.
Two years ago, she made a pivot into machine learning engineeringSpecialised software engineering to design and build artificial intelligence systems that can automate predictive tasks. when she applied for her current job at Cleo, a personal finance app. Her first project involved improving the “quick reply” actions in the company’s chatbot. Watching her work be rolled out for millions of users was incredibly satisfying.
On a typical day, Sneha arrives at the office at 9am. Her team has a morning “standup” to discuss the ticketsIn technology, “tickets” are digital records raised to track and manage tasks or problems. In arboriculture, a common industry term for the specific certificates of competence required to operate machinery such as chainsaws and woodchippers safely. they are working on. Then she spends the rest of the day training models, helping others with their code and perusing the office’s unlimited free snacks.
Data science and machine learning is a well-paid profession. Graduates can expect to earn at least £35k at the start of their career, while seniors will easily earn six figures.
But there can be downsides to working in start-upA new business. or scale-upCompanies that have grown and are no longer in a start-up phase. They are growing quickly and scaling up their operations, hiring new staff and making a profit. tech companies. Some experimental work may never go anywhere, whilst jobs can be at risk as founders search for funding.
Sneha hopes that in five years’ time, she might be a principal machine learning engineer. She is still deciding whether or not she wants to manage people, but in many tech jobs, you can progress without having to manage.
Get creative
Sneha’s top piece of advice is to follow your passions and look for opportunities, from work experience to internships. If you are interested in a tech job, make sure you stick with maths. And keep up to date with the field — AIA computer programme that has been designed to think. has changed the world since she started, and she has to work hard to stay on top of new developments.
Keywords
GCSE – The national exams taken by 15 and 16-year-olds in parts of the UK.
Extended project – An extended project qualification (EPQ) is an optional, self-directed research project that students in the UK can take alongside A-levels. It is worth 50% of an A-level.
UCL – University College London, one of the top universities in the UK.
Algorithms – Any set of rules followed by a computer. In the context of social media, “the algorithm” refers to the intelligent AI that learns the interests of the user and presents them with posts that it thinks will interest them.
Python – A programming language that is often used in web applications, software development, data science and machine learning.
Data scientist – A new type of data expert. Data scientists work with businesses to understand data and help to convert it into action that will improve a business or organisation’s performance.
Education First – Often known as EF, an international education company specialising in language learning.
Natural language processing – A subfield of computer science and AI. It uses machine learning to allow computers to understand and communicate with human language.
Redundant – Not needed any more. Someone who is made redundant loses their job because their work is not needed.
Machine learning – A field of artificial intelligence that aims to use data to teach machines to “learn” for themselves without the need for specific programming.
Machine learning engineering – Specialised software engineering to design and build artificial intelligence systems that can automate predictive tasks.
Tickets – In technology, “tickets” are digital records raised to track and manage tasks or problems. In arboriculture, a common industry term for the specific certificates of competence required to operate machinery such as chainsaws and woodchippers safely.
Start-up – A new business.
Scale-up – Companies that have grown and are no longer in a start-up phase. They are growing quickly and scaling up their operations, hiring new staff and making a profit.
AI – A computer programme that has been designed to think.
