Insights from Data Science Meetup 9.0: AI Coding Assistant
Shivani Makwana

Insights from Data Science Meetup 9.0: AI Coding Assistant

Introduction

Our data science meetup brought together around 40 tech leaders and software engineers looking to discover how AI can be a great coding assistant in the modern software development field.  

Our speaker, Mr. Ashish Kasma, CTO at Lucent Innovation, having expertise and relevant knowledge in AI assistance, presented the practical use cases of coding tools for developers and sparked a conversation that went beyond the hype.  

Let’s explore the key insights and tools discussed at the event and how you can utilize AI for your software development needs.

Unleashing AI in Your Dev Workflows by Ashish Kasma

Mr. Ashish started the event by addressing the fear of every developer: What if AI will take your job? Let’s first come to the fact that AI is no longer a distant concept in development; it is reshaping the way developers write, test, deploy, and maintain code.

But what are the benefits?

Well, there are a lot of benefits of using AI for code generation, such as;  

  • It can reduce boilerplate, allowing developers to focus on logic.  

  • Offers smart suggestions based on the given coding context.  

  • Generates quick prototypes, fostering exploration.  

  • Improves security by detecting risky partners early.

When to Use Coding Assistants?

You can rely on AI coding tools when it's about;  

  • Boilerplate code,  

  • Repetitive tasks,  

  • Speeding up exploration,  

  • Quick fixes & refactors,  

  • Writing tests,  

  • Documentation and comments.

It is a smart decision to rely on an AI assistant, but not every time. Certain situations demand human logic and interaction that AI can’t perform. So refrain from using these tools when;  

  • Solving deep business logic,  

  • Critical security or compliance code,  

  • Refactoring large legacy codebases.  

 Use AI for your job

A Brief on Top AI Coding Tools  

The session continued with further discussion on AI coding tools for developers like Cursor, Copilot, and Trae. Mr. Ashish presented a list of some top tools with their features, ideal use cases, and quick tips as well to help users easily get started, let’s check it out.

1. Copilot  

This tool helps you enhance your everyday coding productivity and handles common tasks with ease. It can be used to write repetitive code, generate unit tests, and learn syntax and APIs.

Features 

  • Real-time code suggestion 

  • Generates code from plain-language prompts 

  • Supports multiple languages 

  • Automatically generates boilerplate, functions, tests, and regex

Limitation 

  • Don’t always understand the deep project context 

  • May generate insecure or incorrect logic 

  • Requires a stable internet connection 

 

2. Cursor  

This AI assistant helps you understand and navigate a large or legacy codebase, writes smarter commits, and helps with debugging and code comprehension.

Features 

  • AI chat window to ask questions  

  • Multi-file context understanding  

  • One-click refactor, debug, and explain code  

  • Quick inline fixes and documentation  

  • Built-in Copilot and comes with advanced features

Limitations 

  • Still being new, some VS Code extensions are not fully supported  

  • Larger than traditional IDEs  

  • Works best when using GPT-4

 

3. Trae 

Trae is a free AI-powered IDE that helps to build quick MVP, provides coding assistance for teams to improve productivity, and automates repetitive coding tasks.

Features 

  • AI-Powered Code Assistance 

  • Builder Mode 

  • Chat Mode 

  • Multimodal Input 

  • Full Codebase Context Analysis 

  • Cross-Platform Support

Limitation  

  • Being a recent tool, you may find compatibility issues with certain extensions and workflows  

  • Internet connection required to utilize AI features powered by GPT-4  

  • While free to use, it is essential to consider data privacy factors while using AI  

 Ashish Kasma on AI

How to Get Started with AI in Your Dev Workflow?  

If you are ready to integrate AI into your development process, check out the following steps to easily get started.   

Start Small: Utilize AI to execute smaller tasks like code generation or documentation.  

Train Your Team: Educate your teams and encourage hands-on practices within teams.  

Integrate Step-by-Step: Add AI tools into your existing IDEs or CI/CD pipelines.  

Analyze and Improve: Lastly, monitor the received output, make necessary changes, and refine usage over time to achieve qualitative outcomes.  

 Data Science Meetup 9.0

Ending Note  

The meetup came to an end with closing thoughts on the fact that AI will not replace developers; instead, it will empower them to build smarter and faster. Our data science meetup 9.0 was filled with thought-provoking discussion and also some viral memes to educate and entertain the audience.

Remember, AI is here to improve productivity and the quality of work, not to take your job. So, embrace it with open hands and give it a try today. That’s it, stay tuned with us for the next data science event.

Also check out: Glimpse from Data Science Meetup 8.0: AI Agents and LLMs