Research Analyst & Data Scientist

Riya Shrestha

Bridging data science and research to uncover insights, tell stories with data, and drive meaningful change

MS + BS UC Berkeley
3+ Years in Research
1 1 Global Markets
Riya Shrestha

Researcher by nature,
data scientist by training.

I'm a data scientist and researcher working at the intersection of human behavior and technology. In my role at BetaHat, I was designing and executing studies for global tech and gaming companies including Microsoft and Riot Games across US, APAC, and EU markets.

I hold an M.S. in Information and Data Science from UC Berkeley, where I also served as Graduate Student Instructor for Psychology and Data Science courses. I'm passionate about applying rigorous research methods to uncover impactful insights from complex, real-world data.

Languages & Tools

PythonRSQLTableauGitAWSSASJavaBashLinux

Libraries & Frameworks

PandasNumPyScikit-learnTensorFlow

Expertise

Mixed-Methods ResearchData ScienceMachine LearningResearch DesignStatistical AnalysisCausal InferenceData EthicsUser ResearchCross-Cultural Studies

Where I've worked

Senior Research Analyst

BetaHat
  • Lead end-to-end market and user experience research projects for major clients (Microsoft, Riot, Zynga, 2K, etc.) independently managing timelines, deliverables, and stakeholder relationships across multiple concurrent studies
  • Design and execute mixed-methods research protocols (surveys, interviews, focus groups) across diverse international populations, ensuring methodological rigor and data quality
  • Synthesize quantitative and qualitative data into comprehensive reports and presentations that inform strategic decision-making for technical and non-technical stakeholders
  • Mentor junior analysts on research design and analytical methodologies while implementing process improvements

Graduate Student Instructor — DATA 104: Human Contexts and Ethics of Data

UC Berkeley
  • Led two 35-person discussions on algorithmic bias, fairness, and responsible AI systems for Berkeley's largest data ethics class
  • Created discussion activities on ethical implications of data science in healthcare and human contexts
  • Graded coursework on applied historical thinking and Science, Technology, and Society (STS) frameworks

Graduate Student Instructor — PSYCH 101: Research and Data Analysis in Psychology

UC Berkeley
  • Guided students in R programming for statistical analysis, including linear regression, ANOVA, and hypothesis testing.
  • Taught curriculum and assisted with coding exercises, debugging, and conceptual understanding of research methodologies.

Research Assistant — East Bay Covid-19 Study

UC Berkeley School of Public Health
  • Supported operations of large-scale public health surveillance study in East Bay Area with thousands of participants
  • Performed statistical analyses in R on survey datasets to explore correlations between demographic factors and COVID-19 prevalence
  • Generated publication-ready charts and tables using R and Microsoft Word; conducted literature reviews on emerging research
  • Served as front-line contact for study participants through email and phone, maintaining clear communication and protocol compliance
  • Recommended improvements to study protocols, participant materials, and communications based on participant feedback

View full work history in CV

Selected work

UC Berkeley MIDS Capstone
96% test accuracy

AudioAid

An edge-deployed AI product for detecting elderly falls through real-time audio classification — enabling offline, privacy-preserving monitoring without cloud dependency.

TensorFlowTensorFlow LitePythonCNNRaspberry Pi
  • Designed, trained, and evaluated a custom CNN to classify fall events from ambient audio using Mel Spectrograms
  • Achieved 96% test accuracy on a high-quality labeled dataset created with volunteers
  • Experimented with transfer learning across multiple model architectures
  • Deployed on Raspberry Pi with live microphone input for offline, real-time monitoring
View project
Data Visualization
50 US States

Cancer Trends in the United States

An interactive data visualization exploring cancer incidence and mortality rates across US states over time, built as a public-facing website.

TableauPython
  • Created a website about cancer incidence/mortality rates over time in the US
  • Developed an interactive heatmap and bar chart race dashboard to track cancer incidence/mortality rates over states and time
UC Berkeley MIDS
n=50 Participants

Information Retention Across Different Mediums

A within-subjects randomized controlled trial evaluating retention of academic material across four media conditions: text, audio, text+audio, and an ADHD-focused extension.

RQualtrics
  • Conducted a within-subjects randomized controlled trial (n=50) to evaluate retention of academic material across four media conditions (text, audio, text+audio, ADHD-focused extension)
  • Used Qualtrics for survey design and data collection; used R for data cleaning, visualization, and inferential analysis including ANOVA modeling and participant attrition adjustment

Academic background

M.S. Information and Data Science

University of California, Berkeley

School of Information

B.S. Molecular Environmental Biology

University of California, Berkeley

Focus: Human Health · Minor in Toxicology

Let's connect.

Whether you have a research collaboration in mind, want to discuss data science, or just want to say hello — I'd love to hear from you.