Imagine when your AI-powered research assistant not only helps you gather and analyze data but also crafts compelling presentations, tracks stakeholder engagement, measures product, revenue, and cultural impact over time, and presents your research share outs for you—all tailored to your unique style and workflow. This isn’t some distant sci-fi future.

I am planning and testing some variations of these activities now for an analogous industry.

Personalized AI agents will fundamentally change the way knowledge workers "work" and conduct research.

As someone deeply embedded in AI, ML, and NLP, I’ve explored these technologies in various client projects researching expectations, limitations, visualizations, and possibilities. A recent study I co-led with over 100 researchers and platform founders, and a current enterprise AI study, further fueled my curiosity how AI tools could become indispensable collaborators in our field. I see a future where a fully customized AI research assistant revolutionizes our research industry.

This article paints a vivid picture of what my personalized AI research assistant could look like and the many implications it holds for our craft. It is based on my first hand experiences conducting research on AI platforms and the people who these platforms are designed to support.

NOTE: The blue content below represents existing AI platform functionality, albeit to varying degrees of success.

The Customized AI Research Assistant: The New Research Partner

Picture an AI assistant, designed specifically for you, which can handle every step of the research process. This assistant could draft research plans, gather and synthesize data, and formulate insights in engaging formats. Imagine the productivity gains and the enhanced focus on strategy and creativity.

My dream AI assistant would seamlessly integrate with tools like Gmail, Google Drive, and Zoom, allowing me to streamline my workflow while maintaining the utmost security and customization. This assistant would learn my preferred style, avoid certain language, and align perfectly with my preferred research recruiting and data gathering approaches. Let’s break down how this might work in practice:

1. Setting Up: Connecting Tools and Securing Data

After logging into my chosen AI platform, I’d configure the assistant to be either public or private and link it with my accounts like Gmail, Google Drive and Zoom. Permissions would ensure that only the necessary files are accessed, maintaining a high level of data security.

2. Teaching the AI: Uploading Key Documents and Establishing a Style Guide

To align with my unique approach, I’d upload essential documents, past presentations, and reference materials. This initial “training” would help the assistant understand my tone, style, and research preferences. It would avoid certain yet common terms I loathe (like “delve” and “realm”) and adhere to specific research preferences, such as use of assumptions over hypotheses in planning phases.

3. Performing Research: Automated Tasks with Precision

Once set up, my assistant could tackle tasks like drafting research plans, summarizing interview transcripts, and categorizing findings by participant segments. It could highlight key quotes, identify recurring themes, and prepare summaries that allow me to quickly compare and contrast different participant groups and substantiate learning with previous research.

NOTE: The following purple content represents what future, multi-modal agent functionality will likely encompass.

And here’s where things get really interesting.

4. Delivering Insights with a Digital Avatar

Imagine your AI assistant not only creates Google Slide decks with brand-consistent templates and royalty-free visuals but also presents your share-outs through a customized digital avatar. This avatar mirrors your style and tone, making the presentation compelling and engaging for stakeholders—even when you’re not physically there.

Your avatar uses storytelling frameworks, like the Hero’s Journey for user narratives or Freytag’s Pyramid to emphasize key findings. It adapts its delivery to fit each audience, extending your reach by presenting simultaneously across multiple teams, to stakeholders in different time zones, and in 7000 languages. These avatars bring 4D life to your research and make lasting impressions, even contributing to research repositories with accessible, dynamic Q&A flywheels that fuel and support future inquiries.

5. Real-World Impact: Multimedia and Engagement Tracking

Taking it a step further, the assistant could generate video highlight reels that are stripped of PII and packed with impactful quotes. Post-presentation, it could track who engages with the presentation, and how, follow up with viewers, and offer actionable recommendations to drive the insights into action across XFN teams.

6. Establishing Baselines, Tracking Impact, and Automated Reporting

Beyond conducting research tasks, my AI assistant would establish and monitor metrics that track the research impact over time:

Setting Baselines for Impact Measurement: Before a project starts, the assistant could analyze historical data and set benchmarks. This might include metrics like user satisfaction or task completion rates, providing a reference point to measure progress and evaluate the effectiveness of implemented changes.
Continuous Impact Tracking and Insights Comparison: As findings are applied, the assistant could monitor real-time data to assess how they influence baseline metrics. It could track engagement rates, (eliminate NPS scores for ever! LOL) or task success among different segments, providing metrics on where improvements have been achieved or where further adjustments are needed. These automatically delivered to the most pertinent Slack channels, tagging the original stakeholders, as well as those new to the project. These metrics also be added to my personal "watch list" which would in turn inform employee performance reviews.
Automated Reporting and Feedback Loops: The assistant would compile regular reports that visualize these metrics, highlight key changes, and provide context, giving stakeholders a clear picture of research impact. Additionally, it could facilitate feedback loops by sending surveys, gathering stakeholder input, and following up quarterly to ensure that what were tracking remain relevant and actionable. This continuous feedback mechanism ensures that your research evolves in line with organizational needs and provides valuable recommendations for future projects.

Future Possibilities: Game-Changing Potential

Here’s where these AI-driven, customized research assistants could take things to the next level:

End-to-End Project Support: Imagine an assistant that not only handles the research process but also anticipates needs, like formatting reports, scheduling meetings, identifying and removing fraudulent participants, freeing you to focus on strategy.
AI-Driven Avatars for Presentations: Picture a digital avatar delivering insights on your behalf, providing a consistent and engaging experience for stakeholders, while saving time and ensuring effective delivery.
Real-Time Engagement Metrics and Follow-Ups: Beyond report distribution, the assistant could monitor who engages with the research, compile summaries, and even automatically follow up to ensure insights are acted upon.

Wrapping Up: New Era Research

The idea of an AI assistant that handles end-to-end research tasks is no longer a distant vision. These assistants exist today and they can evolve into trusted partners, streamlining workflows, enhancing deliverables, and transforming the way insights are identified, shared, and acted upon.

Of course, a skilled researcher remains essential to set up and continuously “coach” the assistant, keeping it aligned with the latest data sources, industry and ethical standards, current events, the competitive landscape, and the many other inputs that go into decision-making.

The agent is not a “set it and forget it” or “auto-pilot” solution—it’s a powerful tool that will thrive under expert guidance and continued nurturing.

So, what would you like your AI research assistant to do for you? What breakthroughs are you most looking forward to in an AI-powered research future?

As AI technology continues to evolve, it will be essential for researchers to stay adept at both leveraging these tools and maintaining their critical role in the research process.

Interestingly, the foundations taught in the Ask Like A Pro series are more important today than when we launched years ago.

Ask Like A Pro On-Demand is now even more relevant and continues to perform well!

Learn more here.

Speak up, get involved, and share the love!

Connect with Michele on LinkedIn for more UXR tips and UX discussions
Read about Curiosity Tank workshops.
interested in team user research training? Let us know!
Participate as an On-Demand, or as an Observer, for less!
Dive into a different UX Research term every week! Sign up for UX Lex weekly emails here.
Forward this email to someone who you think might enjoy it. Better yet, sign them up here. It will be the gift that keeps on giving.

And that’s a wrap!

We try to alternate between a theme and UX/UXR jobs, events, classes, articles, and other happenings every few weeks. Thank you for all of the feedback. Feedback is a gift, and we continue to receive very actionable input on how to make Fuel Your Curiosity more meaningful to you.

What do you think? We're constantly iterating and would love to hear your input.

Stay curious,

- Michele and the Curiosity Tank team

PS: Hit replay to share your thoughts on AI Agents, ask questions about Ask Like A Pro, or just say hi! We'd love to hear from you!

My Customized Research Assistant