I’m a bit late, but better late than never.

Let’s talk agentic AI. AI agents have been around for a while, but we never had the resources to implement them, and make them a part of our work routine.

Thanks to generative AI and LLMs we can.

But what does this mean for data science? One thing is for sure, Agentic AI won’t be replacing us, but here’s what’ll change.

The Role of Data Scientists in Multimodal AI

AI is growing more independent. It can now do certain things autonomously including preprocessing text, images, audio and more.

They can literally generate insights autonomously and help you meet your business goals.

But that doesn’t mean that data scientists are losing control. Instead, they’ll be working on ethics, compliance, guidance, optimization and most importantly ensuring the reliability of the outputs.

How is AI changing Data Science?

I’m focusing now on data science, using both research and how I conduct my work at Aurora Data Strategies. With that in mind, check these:

Giving traditional data analysis superpowers

Manual data processing is becoming a thing of past. That is good because most of the preprocessing is extremely boring and tiring.

With Agentic AI, you can automate these processes and allow for real-time processing and analysis to make its magic. The UAE’s ADNOC uses Agentic AI in the energy sector which allows for faster and more detailed seismic surveys.

Essentially, this helps improve production forecasts.

New Data Roles

The rise of agentic AI is reshaping the role of data scientists. Rather than focusing solely on predefined problems, data scientists are now required to proactively frame complex issues and explore innovative solutions.

This shift calls for a deeper understanding of business contexts and the development of new skill sets, including expertise in agentic AI frameworks and platforms.

Real-time Decision-making

Agentic AI systems can process vast amounts of data in real-time. This is of key importance to decision-makers who need timely and accurate insights.

The autonomous monitoring of data streams continuously, Agentic AI can identify patterns and trends which can benefit different organizations and allow them to make decisions faster.

For example, in healthcare, agentic AI can monitor patient data, adjust treatment recommendations based on new test results, and provide real-time feedback to clinicians, thereby improving patient outcomes.

Iterative Model Development

Agentic AI automates model generation, evaluation, and optimization, continuously refining outputs based on real-time feedback.

Less manual tuning leads to accelerated experimentation. That ultimately means data scientists can focus on strategy rather than constant fine-tuning and evaluation.

We’re talking about better performance and more efficient AI systems.

AI Agents as Research Assistants

Instead of spending hours on grunt work, data scientists will collaborate with AI as co-researchers, accelerating discovery and pushing the boundaries of what’s possible.

Automated Experimentation

Agentic AI is taking experimentation to the next level. Instead of manually setting up tests and tweaking parameters, AI will autonomously design, run, and refine experiments—optimizing architectures, tuning hyperparameters, and selecting the best data augmentations.

What do you think about these? We’re just at the beginning of the year and

Agentic AI is promising so much.

Where do you feel you’ll fit in this setting? Reply to this email or DM me on @codingmermaid.ai

P.S. As AI is growing fast, I’m updating my AI engineer roadmap to have more focus towards Agentic AI.

I added project templates, more project ideas, more resources and a whole new chapter on Agentic AI development.

If you already own the roadmap, all you need to do is refresh or duplicate to your workspace again. If you don’t, you can grab it here!

Become AI-literate

Best, Danica