Pratik Dahal

Ph.D. Candidate in Statistics

University of Arkansas

I work across statistical research, applied data analysis, machine learning, and teaching.

Research Projects Teaching

Selected Projects

Selected applied work across analytics, experimentation, machine learning, and AI systems.

Marketplace Operations Analytics

Operational analytics on a multi-table e-commerce dataset, focused on reliable metrics, data quality, fulfillment performance, seller behavior, and decision-oriented reporting.

  • Integrated relational marketplace data using Python, SQL, and PostgreSQL.
  • Validated joins, referential integrity, metric coverage, and data quality before analysis.
  • Analyzed fulfillment, seller, category, and geographic performance.
  • Developed operational KPIs and dashboard-ready datasets.
  • Python
  • SQL
  • PostgreSQL
  • Tableau

Retail Media Measurement & Decision Platform

Experimentation and measurement on synthetic retail-media data, designed to distinguish attributed performance from true incremental impact using randomized treatment and control groups.

  • Measured incremental conversion, orders, and revenue using randomized holdouts.
  • Estimated uncertainty using confidence intervals and bootstrap methods.
  • Built campaign- and segment-level analytical datasets.
  • Used Python, SQL, and PostgreSQL to support reproducible measurement.
  • Python
  • SQL
  • PostgreSQL
  • Experimentation

Vector RAG Backend

Retrieval-augmented generation backend for retrieving relevant source material and returning grounded responses with citations.

  • Built document ingestion and vector retrieval workflows.
  • Used PostgreSQL and pgvector for retrieval infrastructure.
  • Exposed retrieval and generation through FastAPI.
  • Emphasized source grounding, citations, and evaluation.
  • Python
  • FastAPI
  • PostgreSQL
  • pgvector
  • RAG

Explore all projects →

About

I am a Ph.D. candidate in Statistics at the University of Arkansas. My work spans statistical methodology, applied data analysis, machine learning, and undergraduate teaching.

My doctoral research focuses on statistical models for structured data, while my applied projects explore how statistical and computational methods can be used to investigate practical problems in areas such as marketplace operations, experimentation, search, and AI systems.

Alongside my research and project work, I have taught undergraduate mathematics and statistics, including Principles of Statistics and College Algebra.

Research

My research is in Bayesian computation for structured models: spatial dependence on graphs, variational inference, and MCMC when an approximation needs a check. SDM-CAR is the main example. It puts CAR-type covariance in the graph spectral domain, fits the same filters with collapsed VI or MCMC, and compares them under misspecification and block-missing prediction. The through-line is uncertainty that remains usable when the neighborhood graph, or the observation process, is only partly known.

SDM-CAR code Technical note

Explore Research →

Teaching

I have taught undergraduate mathematics and statistics at the University of Arkansas.

Explore Teaching →

Explore Writing →

Contact

For questions about my work, research, teaching, or professional collaborations, feel free to get in touch.

Email: mapratikdahal@gmail.com

GitHub: github.com/dahlp94