About what I do...

I'm a scientist who somehow ended up loving both genomics and heavy equipment, which says more about how transferable good statistical thinking is than it does about my grand career plan. I've spent the past decade turning messy data into predictive models and productionizing ML systems that people trust to act on, whether the call is how to price a used excavator, which genetic variant is real, or which data to prioritize. These days, I lead pricing analytics and demand forecasting at Caterpillar, where my models help value tens of thousands of machines worth billions of dollars, and forecast industry demand accurately. Before that, I spent nearly seven years at Bayer building genomic data science capability that delivered north of $10M a year in value. I enjoy roles that are part data science and part translation, because what I care about most isn't the algorithm itself, but whether the people making the decision actually understand and trust what the model is telling them.

Experience

• 
  Pricing Analytics & Applied AI Sr. Data Scientist @ CATERPILLAR
  Jun 2025 – Present

  I primarily focus on two areas: leading the development / deployment of ML models for pricing analytics products, and developing time-series models for demand forecasting. My models powers CAT Used Price Advisor, a product used by dealers to value more than 70,000 machines, collectively worth over $12 billion. I expanded the capability of this tool from predicting at region level to predicting at the dealership level that improved local pricing accuracy. I built and own a Used Price Index to track price trend of used machines, which CAT Financial and senior leadership now use to sharpen pricing strategy across different sales channels. This index separates out real market price movement from quality and compositional mix. For my other focus area, I integrated used pricing, fleet condition, aftermarket parts spend, macro economic indicators, and competitive intelligence, and delivered a fifth-quarter forecasting model to predict future industry demand and support resource allocation.

• 
  Applied AI/ML Principal Lead Data Scientist @ BAYER
  Jun 2023 – Jun 2025

  I owned and drove the ML and analytics strategy across a multi-billion dollar product pipeline, pushing prediction accuracy up by 42% and generating more than $10M a year in value. I collaborated with cross-functional teams to productionize models and automate pipelines, supporting business decision-making. I represented crop science & breeding org during planning and rolling out of company-wide GenAI solutions. I hired two data scientists and mentored several on their ML projects and career development, and helped the team stay steady through a reorg.

• 
  Simulation & Modeling Sr. Data Scientist @ BAYER
  Jul 2020 – Jun 2023

  I built a unified genomic feature repository in BigQuery that became the team's single source of truth, and developed an imputation algorithm that took usable data return from 30% to 92% per feature, a big enough jump that it earned me a Top Performance Award. One of my favorite projects was using genomic data from public domain to generate internal resources, which saved the company $500K. In another project, I applied Gaussian mixture model to improve true variant detection, which led to a 70% increase in the number of variants that could be used for downstream analysis.

• 
  Genomics Data Scientist @ BAYER
  Nov 2018 – Jul 2020

  I built an interpolation algorithm and served it via an R Shiny app that cut genomic coordinate conversion time by 90%. I designed hypothesis-driven experiments to characterize genetic variants, the kind of groundwork that makes later discovery work possible. I developed a method with greater statistical power to identify more gene-editing targets. I applied mixed-integer programming to optimize haplotype calling, increasing accuracy by 5% and reducing gaps in genome coverage. I solved a long-standing sample mix-up problem by developing a parent-progeny-trio QC framework, achieving 100% accuracy in identifying mislabeled samples and preventing future errors.

• 
  Doctoral Researcher & Postdoc – Plant Genetics @ KANSAS STATE
  Aug 2013 – Nov 2018

  During my PhD and postdoc, I developed sequencing-based methods to detect genome deletions and mapped insect-resistance genes in wheat, resulting in drought-resistant varieties that need less pesticide. I applied principal component analysis and hierarchical clustering to identify a novel population group, contributing a new source of genetic variation to the wheat research community. I applied mixed-model regression to map insect-resistance loci in the wheat genome, supporting trait discovery and breeding decisions.

• 
  Genomics Data Consultant @ HEARTLAND PLANT INNOVATIONS
  Aug 2017 – Dec 2017

  I provided support for data-driven decision-making by leveraging my expertise in genomics data analysis, collaborating with their researchers to integrate data from a newly acquired business, and making recommendations for selecting germplasm to kickstart their wheat breeding program.

• 
  Graduate Research Assistant – Plant Breeding @ NEW MEXICO STATE
  Aug 2010 – Aug 2013

  Thesis: Selection progress for reduced Iris yellow spot symptom expression in onion.
  I planned field and greenhouse experiments, managed trials, collected data, performed statistical analyses, and advanced experimental lines to the next phase in a classical breeding framework.

Education

• Ph.D., Genetics & Bioinformatics

  Kansas State University · Manhattan, KS

• M.S., Plant Breeding & Statistics

  New Mexico State University · Las Cruces, NM

• B.Sc., Agriculture

  Punjab Agricultural University · Ludhiana, PB, India

Skills

Data Science

Tech Stack

Leadership

Publications

Most of my published work is on genomics and quantitative genetics from my research years.

• 14 scientific publications — see the full list on the Publications page

• Code on GitHub