What are the best AI tools for students to build real-world projects?
by•
Hi,
I am currently in my first year of study as a B.Tech in Computer Science and I am interested in finding out about different types of Artificial Intelligence (AI) based software applications and how they may be used to create real world projects.
Some of the ways that I would like to use AI-based tools are:
To learn coding and programming logic.
To develop web sites, mobile Apps and other AI related projects.
To use these tools to develop innovative ideas and solutions to problems at Hackathons.
To turn an idea into a viable prototype using basic coding skills.
Are there any tools that you have used before that you feel could benefit me? I would particularly appreciate information on inexpensive or free-to-use options.
28 views
Replies
What I build with every day: Claude Code as the coding agent, Flutter for one app on iOS and Android, and Firebase for the backend, where auth, database and hosting have a free tier that's plenty for a student project. But the advice I'd give matters more than the tools: since your goal is learning logic, ask the agent to explain every change before you accept it, and now and then write a small piece yourself and have the AI review it instead of the other way around. And keep a short notes file in the project with the rules you've learned, the agent reads it every time and you'll see your own understanding grow in it.
@gbasovv Thank you so much for the info! I love how you are suggesting having the AI explain each change it makes as opposed to just relying on the AI to generate the new code. I am very much still in the process of learning programming, therefore I do wish to utilize the AI to help me enhance my logic when writing programs but ultimately I will be utilizing it to assist in generating some ideas.
@sara_kottawar1 Glad it helps. One more thing on the ideas part: the agent is great at listing ten ideas and bad at picking one, so pick by what you can finish in a weekend and show to a real person. A small thing that runs beats a big thing half done, and your logic grows from the running one, because that is the one you have to debug.
For free options, try running open models on your own laptop with Ollama. I run Qwen locally on an M1 MacBook Pro, and it costs nothing to use. For coding I mainly use Claude Code, and when it gets stuck I ask Codex, Antigravity or Grok for a second opinion.
At my first job I didn't even know what an OS was, so don't worry about starting from zero. I learned by building.
@hisashispace Thank you again for this information. The fact that you could begin with no prior knowledge and have learned through creating is very encouraging. I will definitely look into both Ollama and local models. As someone who has been working with programming for about 2 months now, would you suggest that I should be spending time learning the fundamentals of python/programming concepts before I try to build small projects while learning (using AI), or should I go ahead and try to build those small projects using AI, and then learn the concepts along the way?
@sara_kottawar1 I'd go ahead and build. That's how I learned too. Early on I only bought programming books that came with sample code, got the samples running, and changed them until they did what I wanted. The concepts made sense once I had something real to hang them on. With AI you can do the same thing much faster, just make sure you ask it why something works, not only to fix it.