Photo by the editor
# Entry
I became interested in artificial intelligence (AI) tools and not only wrote about them, but used them every day in my work as a data analyst. They have completely changed the way I do things, helping me write cleaner code, streamline typing, speed up data analysis, and complete projects much faster.
In this article, I share seven AI tools that have become a enduring part of my workflow. No replacements or substitutes – just the basic building blocks that power everything from machine learning projects to content automation.
# 1. Grammatical artificial intelligence
Photo by the author
Grammatically it’s a tool I’ve been using for almost a decade. It started as a plain grammar and spelling assistant for my assignments and theses, but has evolved into a fully AI-powered writing tool.
Now I can select any text and ask Grammarly to correct it, rewrite it, adjust the tone, or make it clearer, and it always delivers high-quality results.
After Grammarly reviews your content, everything appears sharper, more polished, and ready to publish. I utilize it for my LinkedIn posts, articles, tutorials, project documentation, and emails. It’s one of the few tools I really can’t live without.
# 2.You.com
Photo by the author
I used it Ty.com for two years and honestly, even with the recent subscription price boost, it’s still worth every penny.
I rely on it for research, planning, and learning fresh topics. Deep research mode is one of the best features; researches topics in depth and provides detailed reports that I haven’t seen ChatGPT or other AI assistant.
One of the biggest advantages of You.com is access to top models from Anthropic, OpenAI, Googleand a range of open source models, all in one place. You can test them, compare them and even integrate them into your workflow. What’s more, You.com offers a free Model Context Protocol (MCP) server that makes it incredibly basic to connect your local AI tools and pull results from the Internet in milliseconds.
For work that requires a lot of research or exploring fresh ideas, You.com is definitely one of my most reliable tools.
# 3. Cursor
Photo by the author
I was a fan Cursor long before it became popular. It’s lightweight, intuitive, and one of the first editors to offer native support for agent-based AI workflows.
Today I utilize Cursor z Claude Koda and several key extensions that enable much faster testing, debugging, and shipping of code. I like every detail.
I utilize Cursor to train machine learning models, create websites, build APIs, analyze data, and even put together full projects from scratch. Features like built-in AI suggestions, multi-file inference, instant refactoring, and context-aware planning make you feel like a real AI pair developer.
# 4. Deep note
Photo by the author
Deep note is my favorite tool for rapid prototyping and code testing. I have been using it for five years and it has become a fully capable data science platform. It’s a cloud-based and AI-powered notebook, which means you can simply ask it to analyze your data and it will generate step-by-step code, run it, fix errors, and generate a spotless, structured report for you in the notebook.
It includes knowledgeable autocomplete, debugging support, and a swift loading environment that makes experimenting effortless. I utilize it for my tutorials, demos, and quick experiments, and it has drastically reduced my time creating and testing ideas.
I’ve gotten so used to working in Deepnote that I rarely touch my local notebooks anymore. Everything stays online, organized and synchronized. Nothing beats the work I do.
# 5. Claude’s Code
Photo by the author
Honestly, I was skeptical about Claude Code at first. It seemed too steep to me and didn’t perform well in some of my early data science tests. But everything changed when I discovered that I could integrate GLM coding plan with this. Since then, I have been using Claude Code every day for both personal and work projects.
Using it is hassle-free. I tried open source Twins, Codeand even Droid, but I keep coming back to Claude Code.
Its simplicity, the way it follows instructions and its ability to perform convoluted tasks automatically make it extremely reliable. Nothing else can deliver swift programming, clear reasoning, and support for multi-step coding processes.
# 6. ChatGPT
Photo by the author
How to even start using ChatGPT? It has been a part of my daily life since the day it was released. I utilize it for everything – coding, research, debugging issues, system troubleshooting, writing, and streamlining my workflow.
Whenever I get stuck on a complicated problem, ChatGPT is the first place I turn to for a swift and reliable answer. I ask him personal questions, work-related questions, and everything in between, and he consistently provides useful, context-sensitive answers thanks to his ability to remember past conversations.
What makes ChatGPT so powerful for me is its combination of conversational memory, elastic input, and custom instructions. It adapts to the way I work, understands my patterns, and can switch between tasks effortlessly.
Whether I’m creating code, reviewing notebooks, sketching content, or analyzing data, it’s most like having a full-time AI partner next to me while I work.
# 7. call.cpp
Photo by the author
call.cpp it is the backbone of the local artificial intelligence ecosystem. It is fully open source and allows immense language models to run locally on regular consumer hardware, even without a GPU. It is lithe, swift and extremely competent, providing true bare metal performance. Recently, the developers even added a spotless UI, making it almost feel like a local replacement for ChatGPT.
I utilize llama.cpp for offline projects and any work requiring sensitive code or private data. It easily integrates with local coding agents, chatbots, and custom tools, and setup has become so plain that even Windows users can install it without hassle. Whenever I want to test fresh open source models, I run them directly on my laptop via llama.cpp and share my experience. I also utilize it to generate code, write, and quickly answer questions.
It’s not at the level of ChatGPT, but if you care about privacy, security, and experimenting with fresh models for free, llama.cpp is a tool you want in your stack.
# Final thoughts
My core tools remain the same: Grammarly, You.com, Cursor and ChatGPT. The rest changes depending on my workflow or when better alternatives come along.
As a person with dyslexia, having AI support at their fingertips was a real advantage. These tools lend a hand me understand convoluted text, review my writing, and even tackle research that would normally take me hours to get through. Discovering Grammarly, ChatGPT and Cursor turned what seemed like a challenge into one of my strengths.
I don’t believe that artificial intelligence is here to replace us. It’s here to support us and shape the next generation of workflows where AI becomes a natural part of how we build, write, learn and create. Used well, it doesn’t take away from you; it strengthens them.
Abid Ali Awan (@1abidaliawan) is a certified data science professional who loves building machine learning models. Currently, he focuses on creating content and writing technical blogs about machine learning and data science technologies. Abid holds a Master’s degree in Technology Management and a Bachelor’s degree in Telecommunications Engineering. His vision is to build an AI product using a graph neural network for students struggling with mental illness.