How ChatGPT Boosts My Productivity as a Fullstack Developer

This story is free for everyone. If you cannot access the story, please use this link.

As times are changing and everyone has access to the internet and free knowledge, a lot is expected from an intern or a new joiner, and if you are a senior software developer, it is mandatory to keep up with the ongoing pace and also stay ahead so that you can provide tech guidance when required.

As a senior software engineer, much is expected of me. Beyond managing numerous story points, I also lead a team of three junior developers. My role involves writing code and delivering output that serves as a foundation and example for them to follow in future tasks.

I want to share my story of how AI tools like ChatGPT, CoPilot, and Gemini help me to free up my time so that I can focus on things that matter, i.e., Breaking down complex problems into smaller ones, helping the juniors dev with any blockers, doing analysis and providing quality reviews where required.

Below are some use cases of how ChatGpt has helped me transform my development workflow and the Impact of the same.

Writing Better Code With Test Cases:

When writing code, It is expected to be well-performing and well-tested; it should not raise any P1 or P2, at least when deployed on production.

I’m also responsible for writing test cases and thoroughly documenting my code. This sets a standard for the junior developers on my team, ensuring they follow suit and maintain the same level of quality and diligence in their work.

It takes a lot to do all this within a limited time. And this is where the ChatGPT/Copilot helps.

Example :

Code to find a given element from a collection

The above example is about writing the code to find an element.

If let’s say, sometimes I overlook using the optimal data structure for my code, there could be a null check that I am missing or proper logging is not present. I can ask chat GPT to write the code better and more performant.

I asked the ChatGPT to do the same for the above code, and here is the output:

Code made more performant by CHAT GPT

ChatGPT has often helped me with code suggestions that handled edge cases and improved overall code quality.

Impact :

1. PRs take less time to review.
2. All the new code has already been once reviewed and optimized before a PR is created.

Writing Unit Test Cases :

Test cases are an important part of the development lifecycle, and a well-maintained project must have unit & integration test cases to make life easier; Chat GPT not only helps with writing the unit and integration test cases but also helps in expanding the overall horizon of thought process by generating edge cases as well

Impact :

1. 100% of the newly written code has at least some or all test cases.

Writing Scripts :

As a software developer, you must have encountered a task that requires the same repetitive work every time.

1. It could be changing a tag across 10 files.
2. Doing Analysis of an AB test report.
3. Doing manual deployments. Etc.

If you feel lazy about writing a script for these tasks because of the automation not being prioritized by your manager or you feel too lazy to write a script in an unknown language, CHATGPT can help you with this. All you will have to perform basic testing and navigate GPT to write the clean code.

I have used ChatGPT to generate scripts for all the tasks written above and many more.

Impact :

1. A lot of manual tasks have been automated, saving a lot of story points for the company and saving devs from doing the same old dull manual work.
2. Offline testing, which used to happen on very few test cases, now happens very thoroughly with lots of test cases and in very little time compared to earlier processes.

Writing Crud (Create, Read, Update and Delete) Applications:

Many microservices I have seen in my years of work did not have a UI application, and creating one UI app is usually never prioritized. Every time a change has to be made, it is done through Swagger, Postman, or by directly writing the DB scripts.

Crud apps are generally very simple and don’t require much complex logic to be written; Chat GPT Provides a lot of help with writing simple logic codes.

Asking Q&As:

Sometimes, you are too lazy to google for a task , e.g.

1. writing an elastic search API input to process the text in a certain way.
2. Asking questions about a pre-existing component.

Yes, GPTs do hallucinates sometimes, but if you can a bit of validation on your side, it can save you a lot of time.

Disclaimer : I am allowed to use chat gpt private version and co-pilot in my daily work by my organization.