I built a prototype that gathers information about upcoming and ongoing IPOs, displays the results, and scores each IPO using rules that I created. At the moment, I have to enter a prompt manually whenever I want to refresh the information, which uses a lot of AI tokens and makes daily updates inconvenient.
I would like to turn this into a program or locally running web application on my PC. Ideally, it would fetch IPO information automatically, keep only the fields I care about, apply my predefined scoring system, and display the results in a useful format. I am not asking the AI to invent investment judgments—the scoring methodology and criteria would be provided by me. The goal is mainly to reduce repetitive manual research.
I have limited programming experience. I know basic Python, including functions and object-oriented programming, but programming is not my main field. I expect to use AI to help write code, so I would like guidance on the underlying concepts, tools, and learning path I need to understand well enough to build and maintain this system. What should I learn first, and which technologies would be suitable for a local project like this?
4 Answers
This is best treated as a small data pipeline rather than one big AI application. A practical design would fetch IPO information from a reliable API or other permitted data source, convert it into a consistent format, apply your scoring rules, save the results, and then display them. Start with a command-line Python version before building a web interface. Learn how to make HTTP requests, work with JSON, use SQLite, handle errors, and schedule a daily task. Keep the scoring logic separate from data collection so you can test it with saved examples. AI can generate boilerplate, but you still need to understand the code and errors well enough to verify that the results are correct—especially when making financial decisions.
Before choosing a framework, describe the workflow in plain language: where the IPO data comes from, which fields are required, how missing or conflicting data should be handled, how the score is calculated, and how often the results should refresh. Then divide it into independent pieces such as data retrieval, data cleaning, scoring, storage, and presentation. For each piece, note what you do not understand. Those gaps will give you a focused list of concepts to learn instead of requiring you to study all of programming at once.
Be careful not to treat a working interface as proof that the analysis is accurate. IPO data can be incomplete, delayed, or formatted differently between sources, and a bug in one field could change the score. Keep the original values, record when each update was obtained, log failures, and review the results manually until you trust the pipeline.
Ask an AI coding assistant to build the project in small, explainable steps rather than requesting the entire application in one prompt. Begin with a script that reads a sample JSON or CSV file and calculates scores. Add live data retrieval next, then persistence and scheduling, and only afterward add a local interface. Include tests for every scoring rule and check the output against calculations you perform manually. Also verify that your data provider permits automated access and that the source is reliable enough for your use case.
That makes sense. I will start with the data pipeline and scoring logic instead of trying to build the whole web app at once.