%20(1)%20(1)%20(1).png)
Meet Hans Krebs
Staff Product Designer at DISCO
Creative, avid traveler, and human.
What he does with AI: Designs tools like DISCO’s Cecilia Auto Review that leverage generative AI to enhance the efficiency and accuracy of first-pass document review.
Why it matters: Designers like Hans seek to create easy-to-use tools that remove friction for users doing all kinds of work, no matter the industry.
[fs-toc-h2]Q: As a product designer, how has AI changed the way you approach your work?
Hans Krebs: Although I work with AI and technology, I would still describe myself as a relatively late adopter of generative AI (GenAI). When it comes to new technologies, I tend to hang back and let others identify pain points and explore use cases before I try out later iterations.
Before using any GenAI tool, I take a step back and ask myself a key question: What problem is this actually solving for me? For instance, I’m genuinely intrigued by some of the GenAI image creation tools — they’re undeniably capable of producing impressive results. However, these tools don’t address any specific challenges I encounter in my day-to-day life. For me to invest time and energy into a new technology, it has to clearly solve a real problem I face, rather than being something I feel I should use simply because it exists.
Recently, I encountered a problem that AI could genuinely help me solve. As an avid traveler, I was planning a trip to Vermont and decided to ask a GenAI tool to create an itinerary. It was fascinating to compare the AI-generated suggestions to the itinerary I had already put together — there was quite a bit of overlap! While the AI didn’t change any of my plans, it did reinforce some of the activities I had already prioritized.
This experience made me reflect: If I hadn’t taken the time to create my own itinerary first, would I have simply accepted the AI’s recommendations at face value? Or would I have still felt the need to do my own research?
I think the answer to this question is deeply personal and depends on the individual and the use case. However, the more I use GenAI and see it align with my own findings, the more I find myself starting to trust its outputs.
For me, this process is a great analogy for how I approach AI — and how I suggest other self-proclaimed late adopters approach it, too: trust but verify. Experiment with the technology, try it for specific tasks, compare its results to your own, and refine your inputs before trying again. This iterative approach not only builds confidence but also helps you get the most value out of the tools.
Read more about how to build trust with GenAI tools, including how to allay privacy and security fears: Humans Behind the AI: The Privacy Attorney 🔒
[fs-toc-h2]Q: How have your experiences across diverse industries influenced your approach to design?
Krebs: Before joining DISCO, I worked on AI-powered products at Indeed for several years. The recruiting tool we built utilized machine learning to recommend candidates to employers. As employers reviewed candidates and marked whether they were a good match or not, we fed those signals back into the algorithm to refine its recommendations and deliver more qualified candidates.
This experience was incredibly formative for me. It gave me a deep understanding of how these systems operate and what types of signals are most critical for optimizing the algorithm. It was a constant process of iteration — not just ensuring the tool presented the right candidates to employers, but also improving how it communicated the rationale behind those recommendations. This focus on transparency and usability was key to building trust in the tool and ultimately making it more effective.
It’s actually quite similar to the AI we’ve had at DISCO for many years: predictive tagging. As reviewers tag documents in their database, we feed those signals into the algorithm, enabling the system to learn which tags correspond to which types of documents. Over time, as more documents are tagged, the system becomes increasingly accurate at predicting which tags should be applied to new documents, making the process faster and more efficient.
To learn more about the future of AI in the legal industry, check out: Humans Behind the AI: The Accidental Lawyer ⚖️
[fs-toc-h2]Q: What does it mean to be “user-centered” in your design?
Krebs: No matter the industry — whether it’s legal, HR, gaming, or something else — taking a user-centered approach to designing software means the same thing: clearly understanding who your users are and prioritizing their needs throughout the process. It may sound straightforward, but effective, user-centered design always starts and ends with the user. It’s about uncovering pain points in their current workflows, knowing what they're trying to achieve, and building tools that genuinely meet those needs.
At DISCO, this means constantly seeking feedback. We’re fortunate to have an entire Professional Services team that actively uses the DISCO platform, giving us easy access to firsthand insights from real users. On top of that, we work directly with customers when developing new features, essentially testing our ideas in collaboration with the people they’re meant to serve.
For example, I’m part of the Cecilia Auto Review team, which is a new feature we launched in August 2024. Since this is a fresh capability, we’re working closely with select customers to make optimizations and refine their results. Throughout this process, we’re focused on usability, ensuring both new and experienced users can be productive with the tool.
Ultimately, being "user-centered" means designing tools that solve real problems that users actually experience, not just ones we assume they face. It’s about listening, iterating, and always keeping the human need at the heart of the process.
Related: Smarter, Faster Document Review with Cecilia Auto Review 🔎
[fs-toc-h2]Q: What is your biggest hope and your biggest fear for the future of AI?
Krebs: Before DISCO, I spent a couple of years in the gaming industry, where one of my favorite aspects of the job was working alongside a team of incredibly talented concept artists. They would post their creations on the walls, so walking through the office felt like a journey through imaginative, completely unique characters and environments. It gave me a profound appreciation for human ingenuity and creativity.
One of my concerns is that as AI becomes increasingly adept at emulating human work that it diminishes the value of creativity. I fear that, in time, we might no longer see original drawings on office walls, as AI takes over tasks that were once the result of human inspiration.
I do believe there’s another side to that coin. GenAI also holds enormous potential to enhance and accelerate the creative process, acting as a tool to inspire and streamline our ideas rather than replace them.
What I’m most optimistic about is AI’s potential to handle the truly mundane aspects of our daily tasks. By automating these routine duties, AI can free us to focus on the more meaningful and creative work that we actually want to do.
For further reading on the future of legal technology and how AI is shaping the landscape, click here.
Meet more Humans Behind the AI:
Watch the on demand recording to explore the benefits of tech-driven evidence management.
View more resources
