Tech Roles, Explained

What you actually do all day, in plain English.

Software engineer, ML engineer, data scientist, quant developer… the job titles all sound similar, and job descriptions don’t always help! This guide breaks down six common tech roles for students: what a normal day looks like, the tools you’ll use, and how to start building towards each one right now.

You don’t need to pick one forever. Lots of people move between these roles, and your first internship is the perfect chance to find out what you enjoy!

At a glance

Role	In one line	Most of your time goes on	Best if you love
Software Engineer	Builds and maintains the software people use	Writing, reviewing and testing code	Building things and solving problems step by step
ML Engineer	Gets machine learning models working reliably in real products	Data and model pipelines, deploying and monitoring models	Code AND maths, and seeing models used for real
Data Scientist	Uses data to answer questions and guide decisions	Analysing data, running experiments, presenting findings	Asking “why?” and telling stories with numbers
Quant Developer	Builds the fast, reliable systems behind trading	High-performance code, working with researchers and traders	Maths, speed and high-stakes problems
Product Manager	Decides what gets built and why	Talking to users, planning, working with engineers and designers	People, ideas and the bigger picture
Technology Consultant	Helps organisations choose and roll out technology	Client workshops, analysis, slides and delivering projects	Variety, presenting and seeing lots of industries

Software Engineer (SWE)

In one line: You build and maintain the software that people and businesses use every day, from apps and websites to the systems running behind them.

What you actually do all day

Start with a quick team stand-up: what you did yesterday, what you’re doing today, and anything blocking you.
Write code for a new feature or a bug fix, usually from a ticket on a board like Jira.
Review your teammates’ code and get feedback on yours through pull requests.
Write tests so your change doesn’t break anything else.
Debug! A surprising amount of the job is working out why something isn’t working.
Join planning and design discussions about how to build the next thing.

Quick facts

Tools: Python, Java, JavaScript/TypeScript, C# or Go; Git and GitHub; SQL databases; cloud platforms like AWS, Azure or Google Cloud.
Flavours of SWE: frontend (what users see and click), backend (servers, APIs and databases), full-stack (both!), and platform or DevOps (the systems other engineers build on).
Where you’ll find it: everywhere! Big tech, banks, start-ups, energy companies, retailers, the NHS and government.
You’ll love it if: you enjoy building things, breaking big problems into small steps, and seeing something you made work.

How to start now

Build a project from start to finish and put it on GitHub with a clear README.
Practise data structures and algorithms on LeetCode or HackerRank. Most SWE interviews include them!
Learn Git properly. Every engineering team uses it.

Machine Learning Engineer (MLE)

In one line: You take machine learning models, including LLMs, and make them work reliably inside real products.

What you actually do all day

Prepare and clean the data that models learn from.
Train, fine-tune or evaluate models, and compare them to see which performs best.
Build pipelines so models can be retrained and deployed automatically.
Plug models into products, often through APIs. With LLMs, this includes prompt engineering, retrieval-augmented generation (RAG) and building AI agents.
Monitor models once they’re live: are they still accurate, fast enough and fair?
Write lots of normal software too! ML engineering is software engineering first.

Quick facts

Tools: Python, PyTorch or TensorFlow, scikit-learn, SQL, Docker, cloud ML platforms (AWS SageMaker, Azure ML, Vertex AI) and LLM APIs.
Where you’ll find it: big tech, AI labs and start-ups, and increasingly every industry, from energy and finance to healthcare and retail.
You’ll love it if: you enjoy coding AND maths, and you want to see your models used by real people in real products.

How to start now

Take an ML project beyond the notebook: wrap your model in an API and deploy it.
Get comfortable with the maths behind ML: linear algebra, probability and statistics.
Enter a Kaggle competition or build something with an LLM API.

Data Scientist

In one line: You use data to answer questions and help people make better decisions.

What you actually do all day

Meet stakeholders to understand the question they need answered.
Pull and clean data with SQL and Python. This takes more time than you’d think!
Explore the data, spot patterns, and build statistical or machine learning models.
Design and analyse experiments, like A/B tests.
Build dashboards and visualisations.
Present your findings to non-technical people and recommend what to do next.

Quick facts

Tools: Python (pandas, scikit-learn), SQL, R, Jupyter notebooks, Tableau or Power BI, and Excel.
Where you’ll find it: tech companies, banks, consultancies, retailers, healthcare (including the NHS), government and research.
You’ll love it if: you’re curious, you love asking “why?”, and you enjoy explaining numbers so anyone can understand them.
Good to know: Data Analyst roles are similar but focus more on reporting and dashboards than modelling. They’re a great way in!

How to start now

Pick a public dataset (Kaggle, ONS or data.gov.uk) and use it to answer a real question.
Learn SQL. It’s in nearly every data job description!
Practise presenting: turn one analysis into a short slide deck or blog post.

Quantitative Developer (Quant Dev)

In one line: You build the fast, reliable software that trading firms and banks use to research, price and trade financial products.

What you actually do all day

Work closely with quant researchers, who come up with trading ideas, and with traders.
Turn research ideas into production code that runs trading strategies.
Make systems faster. In some areas, microseconds matter!
Build tools and data pipelines so researchers can test ideas on historical market data.
Keep an eye on live systems during market hours and fix issues quickly.

Quick facts

Tools: C++, Python, Java or C#, Linux, and large time-series datasets.
Where you’ll find it: hedge funds, proprietary trading firms, market makers and investment banks.
You’ll love it if: you enjoy maths, performance and high-stakes problems, and like working in small, fast-moving teams.
Good to know: quant developers build the systems, quant researchers use maths and statistics to find trading strategies, and quant traders make the decisions in the market. Interviews often include maths, probability and puzzles as well as coding.

How to start now

Get strong at one language, ideally C++ or Python, and learn how computers work under the hood (memory and performance).
Practise probability and mental maths. Zetamac is great for speed, and “A Practical Guide to Quantitative Finance Interviews” (the “green book”) is a classic.
Look out for insight weeks at trading firms. Lots of them run programmes for first and second years!

Product Manager (PM)

In one line: You decide what gets built and why, and bring engineers, designers and the business together to make it happen.

What you actually do all day

Talk to users and customers to understand their problems.
Look at data to see how people use the product.
Decide priorities: what the team builds next, and what has to wait.
Write product specs and user stories that explain what to build.
Run meetings with engineers and designers and keep everyone aligned.
Present plans and progress to leadership.

Quick facts

Tools: Jira, Confluence or Notion, Figma (to work with designers), analytics tools like Google Analytics or Amplitude, SQL (a big plus!) and lots of slides.
Where you’ll find it: tech companies, start-ups, banks and any company with digital products. Graduate schemes are often called Associate Product Manager (APM) programmes.
You’ll love it if: you love people and ideas, you can explain things clearly, and you like thinking about the bigger picture. You won’t write code every day, but understanding how software is built helps a lot!

How to start now

Pick an app you use and write a one-page doc: what problem would you fix, and how would you measure success?
Lead a project in a society or a hackathon team. Product management is all about coordinating people!
Learn basic SQL and how software teams work (Agile and sprints).

Technology Consultant

In one line: You help organisations work out which technology they need, then help them put it in place.

What you actually do all day

Work on client projects, often switching industries every few months.
Run workshops with clients to understand their problems and processes.
Analyse data and processes, then recommend solutions.
Help roll out systems like cloud platforms, Salesforce or SAP, or support AI, data and cyber security projects.
Build LOTS of PowerPoint slides and Excel models!
Present to clients, sometimes at their offices.

Quick facts

Tools: PowerPoint, Excel and project management tools, plus cloud platforms, CRM systems, data tools or code depending on the project.
Where you’ll find it: the Big 4 (Deloitte, PwC, EY and KPMG), Accenture, IBM, Capgemini and specialist consultancies.
You’ll love it if: you like variety, enjoy working with people and presenting, and want to see lots of different industries early in your career.

How to start now

Practise case-style problem solving: breaking a messy problem into clear, logical parts.
Get comfortable presenting. Societies, competitions and hackathon pitches all help!
Follow tech news so you understand what businesses are trying to solve, like AI adoption, cloud and cyber security.

Other roles worth knowing about

Data Engineer: builds the pipelines and databases that data scientists and ML engineers rely on.
Cyber Security: protects systems and data from attacks, through penetration testing, security operations and risk work.
DevOps / Site Reliability Engineer (SRE): keeps systems running smoothly and automates how software is released.
UX/UI Designer: designs how products look and feel, based on research with real users.
Business / Technology Analyst: turns what business teams need into clear requirements for engineers.

Still not sure? Ask yourself…

Do I love building things? Try software engineering.
Do I love maths AND code? Look at ML engineering or quant development.
Do I love asking questions about data? Data science or data analytics could be for you.
Do I love people, ideas and planning? Have a look at product management.
Do I love variety and presenting? Technology consulting might be your thing!

How to find out for sure

Do insight days, spring weeks and virtual experience programmes (e.g. on Forage). They exist for exactly this!
Message people in these roles on LinkedIn and ask what their day looks like. Most people are happy to help a student!
Try a small project in each area: a website, a data analysis, a simple ML model.
Remember, your first role isn’t forever! Plenty of people start in software engineering and move into ML, product or consulting later.

More free resources coming soon. Follow @mobsintech for updates!