About

How I think and what I build

I am Alex Kurkar, a Computer Science student at the University of Nottingham. I am drawn to the parts of computing where a small set of rules turns into a working system: parsers, runtimes, interpreters, virtual machines, and the machinery beneath programming languages.

My favourite projects usually start with the same question: what is really happening underneath? That has led me toward compilers and language design, but also toward mathematics, operating systems, machine learning systems, competitive programming, and building tools without hiding too much behind frameworks.

You can also read the compact version here: Alex Kurkar CV.

Education

  • University of Nottingham BSc Computer Science, Sept 2025-Jun 2028.
  • Altrincham Grammar School for Boys A-levels: Mathematics A*, Physics A, Further Maths A, Computer Science A.

Current Focus

  • Language implementation Hand-written lexers, parsers, AST evaluation, bytecode, and stack-based virtual machines.
  • Systems foundations C, C++, Linux, ARM, memory models, and the practical details that make software feel real.
  • Mathematical thinking The algebraic and logical structure behind programming languages and computation.
  • Applied ML and data systems From-scratch neural networks, model evaluation, structured data pipelines, and pragmatic AI tooling.

Experience

  • Code Computerlove Product Development Intern, prototyping a consumer-facing application with user-centred UI/UX, Figma, full-stack implementation, and agile sprint practice.
  • Cheshire Datasystems Limited Software Engineering Intern, learning large-scale system architecture, implementation, QA workflows, and reliable delivery from requirements through testing and deployment.

Technical Range

I work across Python, C, C#, Java, JavaScript, and HTML/CSS, with practical experience in React, Next.js, FastAPI, Flask, Node.js, Docker, Git, NumPy, pandas, scikit-learn, TensorFlow, and PyTorch.

Working Style

I like projects that make me earn the abstraction. I will use libraries when they are the right tool, but I learn best by building the core idea myself first: token by token, instruction by instruction, and test by test.