My favorite AI tools for Data Analysts

*Not sponsored — just my honest recommendations!

AI is going to be a critical part of any job in 2026 and beyond. But at the same time, it feels like there are a million AI tools that being released every day. So I did some hands-on investigating to come up with a list of AI tools that are actually useful.

These are my opinions as of December 2025. I do expect that this list & ranking will change as more tools get released and models get better.

Alright let’s jump right into it.

Photo by Andrea De Santis on Unsplash

Best for coding: Claude Code

As of December 2025, you need the Claude Pro ($20 per month) plan to use it.

My favorite tool for agentic coding. This is the only tool that I trust to autonomously write extensive code for me. It also seems to have gotten even better with Opus 4.5. It does a decent job with coming with analytics + ML Python code, and is able to interpret context around the code pretty well. But even then, I always check my code carefully before I commit any changes.

Three tips for working with Claude Code (that have helped me a lot):

1. Give Claude Code a lot of context on the business, the team and the data. Basically, try to take what is in your head and give it to Claude Code.

2. Use claude.md file: this is the file that Claude Code automatically pulls into context when starting a conversation. You can then copy-and-paste this across different projects.

3. Use their plan mode to align on an analysis plan before you allow it to start building.

Best coding assistant for Python notebooks: Cursor (temporarily)

As of December 2025, you can use Cursor for free, but with usage limits. You’ll probably end up needing to upgrade to their $20 per month plan.

This section is is tough because I really couldn’t a product that works as seamlessly as Claude Code (in the previous section). I’ve tested a few different tools in-depth: Google Colab, Cursor, Jupyter Notebook + AI, Sphinx etc. For most of these tools, I felt like they added more hassle than were helpful. I ended up having to go back and delete a lot of code.

So for now, I’ll give the #1 spot to Cursor. You can use their regular AI chat + agent interface with an .ipynb and it is able to read the cells and write into cells.

Small caveat, the experience is a little buggy right now so it takes some time getting use to the quirks (like the notebook randomly jumping a different spot when a cell is run, or AI agent not being able to write into the cells at times 🤷🏻‍♀️)

Best for automating analytics processes: n8n

As of December 2025, you can self-host your n8n instance for free, or get a cloud-hosted instance for $20 per month.

Anyway, n8n is a AI workflow automation tool. They have both a visual editor and an option to code. I personally use a combination of both — depending on my use case.

• I use the visual editor to make small updates, or when I’m working with nodes that I haven’t used before.

• I use the code option in conjunction with Claude Code (lol.. again I am a fan!) for building out processes from scratch. This helps reduce start-up time to just 5 minutes, before I go back into the visual editor.

I also like that they have AI nodes, that they seem to keep building out. You can easily query the big AI models like OpenIA, Anthropic and Gemini. Or you can get into more sophisticated options, like their AI agent nodes.

Btw, I love n8n so much that I have a LinkedIn Learning course around n8n for Data Analytics coming up soon. Stay tuned!

Best for creating presentations: Gamma

As of December 2025, you can start for free, but you’ll probably have to end up upgrading to their $8 per month plan.

Admittedly, I haven’t spend much time in this space, because I rarely create decks these day. But so far, my go-to for creating slide decks has been Gamma.

I really like the basic templates that you can start with, and how you can use AI to create new templates, that are unique to YOU. However, I would say, I’ve been in some meetings where the presenter is presently directly from Gamma and ran into some lagging issues. Nothing bad, but worth a note if you’re an anxious presentor!

Also, they’ve expanded into documents and websites too (I have not played with this but will give this a shot!)

Best for developing light weight apps: Replit

As of December 2025, you can start for free, but you’ll definitely have to end up upgrading to their $25 per month plan.

I like Replit the best among the agentic, no-code app builder platforms (Lovable and Base44 being the other two that I hear about most frequently).

I find that Replit is able to follow my instructions most closely. It’s simple but the most important criteria for using AI agents. Also, I really like how Replit is self-aware, which means you can ask it questions about Replit itself. I find that to experience to be really delightful!

But maybe a more important question is what kind of lightweight apps might you build as a Daa Analyst? Some ideas to get your brain juice flowing

• A lightweight dashboard with your team’s metrics

• An ad performance prediction calculator

• Your own fully customizable Kanban board

Alright, this is it for my list of favorite AI tools for Data Analysts (for now at least). I feel like there is so much more in this space that I want to explore. If you have any tool recommendations, please email me or comment below in Substack!

ICYMI (in case you missed it!)

• I have a free SQL challenge running in December 2025. It’s super fun, and 2000+ people have already joined! Join the challenge & level up your SQL skills here!

• I am making daily videos breaking down the challenge question (in painful detail) on my Instagram. Check out one of the hardest SQL breakdowns I’ve done.

• This founder story that I am absolutely obsessed with — Jack Zhang turned down a $1.2 BILLION investment from Stripe, and his company now worth $8 billion.

• Difference between a Professional Data Scientist and Forever Aspiring Data Scientist

• Exploratory Data Analysis in SQL cheatsheet

• Data Cleaning in Python cheatsheet

---

By the way, I have also recently started to use Oxylabs.io AI Studio for scraping data! They are a low-code platform where you describe what you need (in plain English), and the AI does the crawling, parsing, and extraction for you. 😲

They have

• AI-Scraper: Extracts structured data from any webpage

• Browser Agent: Controls a browser, replicating human browsing behavior (clicking, scrolling, navigating)

• AI-Crawler: Finds the relevant pages on a site, and extracts data from them

• AI-Search: Performs a web search and information consolidation for you

• AI-Map: Maps an entire domain, surfaces important URLs, and shows you the structure

Whether you’re technical or not, you can now scrape data using plain English with Oxylabs AI Studio → https://oxylabs.go2cloud.org/aff_c?offer_id=7&aff_id=1644&url_id=176