Timing: 2024

Scale: 3 months

The Problem

Generative AI as an emerging product domain

In May 2024, Dell Technology revealed a strategic alliance with Nvidia, along with a new series of products targeted at customers interested in deploying Generative AI (GenAI). Traditionally, Dell's focus was primarily on hardware and infrastructure; however, the integration of GenAI has shifted the focus towards understanding the business outcomes that customers aim to achieve.

As sales and agreements progressed, it became apparent that only a limited number of employees were well-versed in this domain and the specific needs of customers. As a researcher, I took the initiative to identify and consult with various internal subject matter experts to deepen my understanding of how Dell utilizes GenAI internally and to explore the types of challenges Dell is currently helping customers address.

Approach #1: Subject Matter Interviews

Within the extensive framework of our large organization, I diligently explored numerous internal sites and presentations, and reached out to colleagues to locate experts knowledgeable in Generative AI. I soon discovered that expertise within Dell was distributed across several specialized groups:

• A strong internal team that has developed a data science and AI platform to support its internal data scientists, developers, and machine learning engineers.

• A group of technical experts who continuously assess new tools and technologies in Generative AI, sharing their findings with the broader team

• Sales and Consulting team members who maintain direct interactions with customers

I connected with subject matter experts from different areas to gather their varied insights. Through these interactions, I developed a thorough understanding of how Dell utilizes Generative AI to enhance business processes and existing products. I also learned about the diverse customer base that our Sales and Consulting teams support and the specific requirements these customers approach Dell with.

Impact from Subject Matter Interviews

Connect siloed teams

As I engaged with the various groups, I noticed that although they held valuable information, it was not widely shared with the Product teams. To bridge this gap, I facilitated a connection between the main Product team, focused on AI solutions, and the sales team. This helped the Product team identify and understand discrepancies between the product's offerings and its real-world applications. Additionally, I linked the Product team with the Consulting team to initiate discussions aimed at minimizing redundant efforts and exploring ways to productize solutions that the Consulting team had already developed for customers.

Build trust with Product Management

Due to the connections I established, I was able to add significant value to the Product team even before officially joining their AI solutions project. The team, primarily focused on the technical facets of the AI solution—such as necessary technology, vendor partnerships, feasibility, and timelines—was highly receptive to the insights and customer stories I gathered. Eager to delve deeper, the Product team showed a strong interest in conducting in-depth research to understand how customers intend to use Generative AI for business outcomes and how these implementations might vary.

Define workflows and requirements using customer stories

A different Product team developing an AI assistant tool approached the UX team for screen designs that shows conversations between the customer and the AI. Despite several rounds of clarifying requirements, the team struggled to articulate why customers needed this tool, what detailed informations the tool can provide, and how it integrated into existing workflows. Utilizing insights I had gained from the Consulting team about the type of customers they interact with, and how they helped the customers identify business use cases that can benefit from AI, I worked with the UX Designer to create a storyboard. This approach helped to clarify the workflows and requirements, and facilitated more effective alignment with both the Product and Engineering teams.

Approach #2: In-depth Interviews

Purposeful selection of participants

The AI solution the Product team is developing is targeted at customers seeking support and guidance with GenAI deployments. Given the nascent state of this product domain, I recommended examining leaders who are already implementing GenAI in production rather than focusing solely on our direct target market. The objective is to understand the strategies these leaders use, the challenges they encounter, and how they overcome them. This knowledge will enable Dell to adopt best practices and create products that address similar challenges to better support customers who require assistance.

I conducted 10 in-depth interviews with GenAI architects from various enterprise companies, exploring both cloud and on-premise deployments to gather insights into how Dell, which primarily focuses on on-premise GenAI, can enhance its role in the ecosystem. Additionally, I held 3 sessions with consultants to gain a broader understanding of the customer types, use cases, and challenges encountered in the field.

Research questions crafted with outcome in mind

I collaborated closely with the Product team to determine the specific types of information they needed and how this information would influence their decisions regarding features and priorities. This understanding allowed me to develop a tailored discussion guide and effectively pivot my questions during interviews to explore topics of interest raised by the participants. The proactive alignment ensured that the Product team was supportive when I adapted questions in real-time to better suit the technical maturity and business context of each participant—for example, shifting from asking “How would you like a vendor to help you do X?“ to "How are you currently doing X?"

The Product team actively participated in all 13 interview sessions, observing and interjecting with additional technical questions related to GenAI implementation that piqued their interest. This collaborative approach enriched the dialogue and deepened our collective understanding of customer needs and challenges.

Synthesize with maturity in mind

The interviews revealed distinct approaches and challenges among customers, varying according to their levels of data, technical, and process maturity. Beyond gathering the specific information requested by the Product team, such as the list of Large Language Models (LLMs) customers are deploying and their GenAI use cases, I developed a maturity framework to classify customers based on these dimensions. Additionally, I highlighted how different maturity levels influence other technical considerations of interest to the team. For instance, more technically mature customers often plan integration into their existing enterprise environments before completing a Proof-of-Concept, rather than waiting until after. This analysis helped the Product team contextualize the spectrum of customer categories, along with their corresponding needs and challenges.

Impact from In-depth Interviews

Product and services opportunities identified

The Consulting team had been working with a maturity model that aligned with the levels and definitions I synthesized from my interviews. These findings offered a nuanced perspective that helps Consulting teams identify development areas essential for supporting customers with varying maturity levels.

Meanwhile, the Product team gained valuable real-world insights on how leaders in GenAI navigate the technology, team dynamics, and processes. This information is crucial for prioritizing GenAI outcomes to be integrated into our products. Additionally, the research highlighted potential opportunities for integrating additional data management and MLOps into our AI offerings.

Higher UX visibility

Throughout the research process, the Product team was highly engaged and proactive. They appreciated how the customer-focused research complemented their existing technical efforts and were keen to further explore different user personas. This enhanced collaboration fostered a better relationship between UX and Product Management, leading to more open and reciprocal exchange of ideas on later design efforts.

Educated other UX team members on GenAI

Having spent three months immersed in the GenAI field, I synthesized the information I gathered from various internal subject matter experts and from my own research into a comprehensive "GenAI 101" presentation for the broader UX team. This presentation expanded on the foundational GenAI training that all employees have completed. It specifically aimed to help UX team members understand the practical applications of GenAI, its benefits to different business processes, how Dell is incorporating GenAI into its operations, and crucially, the GenAI-related products Dell is developing. This targeted educational effort was designed to streamline the onboarding process for UX team members who may be assigned to new projects, ensuring they have a solid understanding of the context and expectations.