I'm studying AI and machine learning and currently know SQL, C, Java, and Python, including NumPy, pandas, Matplotlib, and some computer vision. I also have basic experience with HTML, CSS, and JavaScript. Since AI/ML roles can be limited, I'm unsure whether to focus on an AI/ML career or build toward web development for better future opportunities. What technical skills, tools, and technologies should I learn next?
3 Answers
If you’re pursuing actual machine learning rather than only connecting existing AI APIs, don’t overlook the fundamentals. Statistics, linear algebra, numerical methods, optimization, and model evaluation will matter more than any particular framework. At the same time, you can keep your career options open by learning software engineering basics: clean Python, SQL, testing, APIs, Git, Docker, and cloud deployment. You don’t have to choose between AI and web development immediately—build AI projects with simple web interfaces and see which parts of the work you enjoy.
You already have a solid foundation. I’d focus on PyTorch and Hugging Face alongside your computer-vision experience, since language and LLM-related work is common in current AI applications. The biggest portfolio gap for many learners is deployment: build a model, expose it through FastAPI, containerize it with Docker, and try a vector database such as pgvector. You don’t need to prioritize C or Java right now unless a specific job requires them. Demonstrating that you can take a model from training to a usable service will make your skills much more practical.
Your next step depends partly on the type of role you want, but a strong general AI path would be PyTorch, scikit-learn, Git, and GitHub. After that, learn how to serve models with FastAPI, package them with Docker, and deploy at least one project to AWS or GCP. Choose one direction rather than collecting frameworks: computer vision with YOLO and OpenCV, NLP/LLMs with Hugging Face and retrieval-augmented generation, or general ML engineering. Your web knowledge is useful too, so learning basic React could help you build complete AI products. One working, deployed project is usually more valuable than several unfinished courses.
I mean actual machine learning. I’m studying for a degree in AI/ML, but there don’t seem to be as many dedicated roles, so I’m considering whether to target AI/ML or web development for better long-term opportunities.