MIT-led case studies exploring clustering, regression, classification, and dimensionality reduction—applied to real-world business problems.
Addressing the new
public health constraints
of a global viral pandemic with
an academically driven public
service announcement
Date Completed
April, 2025
Overview
These case studies were completed as part of the MIT No-Code AI and Machine Learning program. Each project focuses on a specific algorithmic technique—such as classification, clustering, regression, forecasting, or dimensionality reduction—applied to real-world business problems across industries including transportation, retail, sports, and media.
While these aren’t UX case studies, they build the technical foundation for applying AI and machine learning in product design roles. This training equips designers to work more effectively in data-driven environments—where framing the right questions, understanding model outputs, and collaborating with data scientists and engineers are essential to building smarter, more responsive products.
These methods can support product teams by:
- Analyzing unstructured user feedback from interviews or workshops
- Segmenting behavior to drive personalized experiences
- Modeling engagement or churn based on analytics
- Partnering on algorithm-driven features, such as recommendation engines or predictive UI components
This series documents my hands-on exploration of these techniques—developing fluency in the workflows that increasingly power how modern product experiences are designed, validated, and optimized.
Transforming Industries through Data Science
Understanding the Role of AI and Data Science in Business: A Product Designer’s Perspective
This video, Transforming Industries Using Data Science, serves as a foundational module in the MIT Professional Education course on No-Code AI and Machine Learning. It introduces learners to the concrete ways AI and machine learning are transforming businesses—not as abstract technologies, but as tools that shape customer experience, operational efficiency, and strategic decision-making across industries. As a senior product designer, I’m sharing this here to demonstrate how gaining fluency in these applied AI contexts enables designers to more effectively architect the experiences in which these technologies live.
This video, Transforming Industries Using Data Science, serves as a foundational module in the MIT Professional Education course on No-Code AI and Machine Learning. It introduces learners to the concrete ways AI and machine learning are transforming businesses—not as abstract technologies, but as tools that shape customer experience, operational efficiency, and strategic decision-making across industries. As a senior product designer, I’m sharing this here to demonstrate how gaining fluency in these applied AI contexts enables designers to more effectively architect the experiences in which these technologies live.
The course emphasizes that machine learning is most effective when deeply contextualized within a domain—whether retail, healthcare, or finance. Data science techniques alone (statistical modeling, clustering, predictive algorithms) can generate models, but to drive impact, one must understand the workflows, constraints, and goals specific to a business function. This positions designers not just as interface makers, but as partners in defining problem spaces and shaping data-informed interventions. Several concrete business domains are explored in depth, illustrating how AI and data science are applied to real-world challenges. These include:
- Retail Inventory Optimization: Using time-series forecasting, clustering, and optimization techniques, businesses can align inventory levels with demand patterns and supply chain constraints. These insights must be surfaced through dashboards and decision tools that are usable, timely, and tailored to business roles—interfaces designers play a key role in shaping.
- Marketing and Personalization: Segmentation models and recommendation systems help personalize offers and experiences. Designers must ensure that these data-driven interactions feel natural and helpful—whether they appear as curated product suggestions, dynamic content, or personalized messaging flows.
- Healthcare Risk and Diagnosis: Clustering and predictive models are used to assess disease risk, personalize interventions, and optimize resource allocation. Designers can translate complex clinical insights into patient-facing dashboards or decision support tools that improve access and understanding.
- Finance and Fraud Detection: Classification models and decision trees help detect anomalous transactions and flag potential fraud. The design challenge is to surface these alerts meaningfully—without overwhelming users or damaging trust—and support efficient resolution workflows.
- Supply Chain and Demand Forecasting: In domains like logistics and fulfillment, predictive and optimization models help reduce delays, manage costs, and maintain service levels. Designers contribute by building interfaces that allow for scenario modeling, SLA monitoring, and timely course correction.
- Policy and Compliance: Particularly in healthcare and banking, AI is used to identify regulatory risks, assess systemic exposure, and shape policy. These tools often require thoughtful UI patterns for interpreting uncertainty, surfacing recommendations, and maintaining auditability.
- Customer Engagement and Journey Optimization: Businesses track behavior across channels to optimize the path from discovery to purchase. AI helps identify friction points and opportunities for conversion, but the user experience must be orchestrated seamlessly—whether on the web, in-store, or in-app.
These examples show how domain understanding drives both the modeling choices made by data scientists and the design decisions made by product teams. The systems we design are not just visual containers—they are mediators between raw intelligence and real human impact.
Throughout the video, the instructor emphasizes that problem definition is foundational. Businesses must clearly understand their current state, articulate the desired future state, and pose the right questions. In this way, data science mirrors design thinking: both are processes for framing the right problems before building the right solutions.
By embedding this content alongside my portfolio, I aim to clarify not just that I’m learning AI—but how I’m learning to apply it. I’m gaining the strategic fluency to design the experiences where machine learning models live—whether embedded in dashboards, customer journeys, or internal tools. As AI becomes more deeply embedded in enterprise software, this perspective becomes not just useful, but essential.