Since launching Riley, we’ve heard from countless researchers and product leaders about how AI is transforming their work.
A senior user researcher told us, “I shared insights from Riley and finally convinced our exec team on a new product direction.” Another product manager shared, “Riley helped me uncover valuable insights quickly—now I want my whole org using it.”
I’ve been building AI models since 2011 and have watched them grow more powerful over the decade. Technologies like Generative Adversarial Networks (GANs) and Word2Vec laid the groundwork for the rapid advancements we see today. I remember a friend forwarding me a hot new paper called Attention is All You Need in 2017, back when I was still working as a data scientist. Little did I know, I was witnessing the birth of transformer architecture, the technology that would go on to revolutionize NLP and even computer vision.
Our goal with Riley is to make AI deeply accessible and useful for researchers, product managers, and insights leaders. But I also want to do this responsibly and ethically - because at the end of the day, why build something if it’s not genuinely useful?
That’s exactly why Riley is built to work with researchers and product managers, not replace them. In this essay, I’ll break down how AI serves as a force multiplier for Product Insights, enabling you to move faster, uncover deeper patterns, and make more confident decisions.
When it comes to making sense of data, AI isn’t here to replace insights practitioners - it’s here to do the heavy lifting so you can focus on what really matters.
Here is where AI shines:
Bringing Data Together: AI pulls in qualitative feedback, analytics, surveys, and other data sources into one place, making it easier to see the full picture (Silva, Machado, & Silva, 2020).
Spotting Hidden Patterns: Machine learning surfaces trends and correlations that might go unnoticed otherwise. The best insights come from models trained specifically for your organization, evolving with your business. Off-the-shelf LLMs can be helpful, but it lacks the deep context needed to help you accurately distill the insights that are most relevant and current to your organization (Fawaz, Forestier, & Weber, 2019).
Automating the Busywork: AI-powered tools classify qualitative data, summarize findings, and standardize reporting. This frees our time up for more strategic thinking (Anwar, Shukla, & Ali, 2020).
By handling tedious tasks, AI lets researchers, product managers, and insight leaders focus on decision-making, storytelling, and faster execution of strategy (Stoklosa, McKenzie, & Lee, 2023).
From my exploration, here are three areas where AI is currently lacking. Note that everyone building AI, including us here at Riley, are making models smarter and better each day - so this list is a snapshot in time and will likely change soon.
Understanding Complex Contexts: AI struggles to fully grasp the nuances of specific organizational contexts, customer personas, and industry dynamics that are critical for accurate insights generation. For instance, in qualitative research, AI may miss cultural context or emotional triggers that significantly impact consumer decisions, leading to superficial insights (Ribeiro et al., 2020).
This statement was shared by a customer during a usability study:
”Oh great, yet another update!”
I ran this sentiment against 2 common sentiment models:
This is the result:
As you can see, both models categorized the statement as “positive”, whereas, a researcher might detect underlying frustration in a sarcastic comment, and correct this classification as negative. This additional depth of interpretation, especially if the data was gathered across different cultures with even more specific nuances to consider, allows researchers to uncover underlying pain points and prioritize the right fixes.
Inability to Provide Contextual Judgment: AI struggles to make judgments based on broader context, often missing crucial factors that influence product insights. For instance, while AI can identify patterns in user feedback, it may not understand the historical or market context behind them, something human practitioners easily recognize (Silva et al., 2020). However, with adaptive AI - like what we’re building at Riley - this limitation is quickly changing. Our models are evolving to incorporate more context, improving their ability to interpret data with greater accuracy and relevance in real-time.
Adaptability to Evolving Requirements: AI systems often lack the flexibility to adapt to changing design requirements or evolving research needs. They follow predefined algorithms, which can lead to rigidity in their approach and outputs (Tenenbaum et al., 2021).
The last two points show that as AI models improve to incorporate better context, the people who benefit most are those who have been using them from the beginning and have a wider context window (more data that you’ve trained the AI on to make better recommendations over time). This is why I always encourage organizations to incorporate AI into their practice starting today!
Since our goal is to create AI that is responsible, deeply useful, and valuable for insights practitioners, such as researchers and product managers.
Here’s where we believe AI can provide the most value today:
Polishing the Process: Automating Tedious Tasks
Data transcription & analysis: AI can transcribe interviews, highlight key themes, and organize qualitative data quickly.
Survey analysis: AI can analyze large-scale surveys and identify trends in responses faster than manual efforts.
Revealing Hidden Facets: Enhancing Pattern Recognition
AI can surface hidden trends in user data, providing insights that might take humans longer to uncover. Machine learning models can cluster user behaviors into meaningful groups, helping segment audiences more effectively.
Cutting Through the Noise: Generating Hypotheses
AI can suggest potential research directions based on data patterns, helping researchers refine their focus. It can simulate different user flows and predict friction points, guiding UX teams to prioritize areas of improvement.
Sharpening Impact: Distributing Insights Effectively
The most impactful research and product teams I’ve been part of are the ones that can distribute the right insights at the right time to the right stakeholders. When I figured out how to achieve this at a previous company, it made a huge difference and was one of the biggest reasons we were able to drive over $30M in revenue in just one year. AI can streamline this process, ensuring that the right insights are distilled and communicated in a way that is most impactful to the people who need to hear them.
We’re building Riley as your personalized product insights assistant to help you become a super researcher!
The rise of the super researcher isn’t about AI replacing humans - it’s about sharpening our edge. By handling repetitive tasks like data wrangling, analysis, and pattern recognition, AI lets researchers and product managers focus on what truly matters: fast, strategic decision-making and continuous innovation. In this new era, AI doesn’t just augment research; it amplifies human expertise, making us shine brighter and become more capable than ever.
Want to see how Riley can help you become a super researcher? Try for free here
Claudia is the CEO & Co-Founder of Riley AI. Prior to founding Riley AI, Claudia led product, research, and data science teams across the Enterprise and Financial Technology space. Her product strategies led to a $5B total valuation, a successful international acquisition, and scaled organizations to multi-million dollars in revenue. Claudia is passionate about making data-driven strategies collaborative and accessible to every single organization.
Claudia completed her MBA and Bachelor degrees at the University of California, Berkeley.
LinkedIn