ML Specialist - Regression Modeling and LLM

About Cherry

We are product team. Cherry is a data-driven marketplace for commercial vehicles — including trucks, trailers, and heavy equipment.
One of our core products, the Cherry Market Rate (CMR), is a valuation engine that predicts wholesale and retail prices based on specifications like mileage, age, and engine type.

We’re expanding the CMR pipeline and looking for a Machine Learning Specialist with experience in regression modeling, data scraping, and LLM integration to strengthen our data accuracy and build intelligent automation tools for marketplace users.

What You’ll Work On

You will collaborate with our backend and data engineering teams to enhance:

• Cherry Market Rate (CMR) regression pipelines — improving accuracy and stability of vehicle price predictions.
• Automated data scraping for sourcing real market listings
• LLM-based listing assistant that helps users create listings from plain text input.

Responsibilities

• Analyze and optimize the current regression-based valuation model (isotonic regression, IQR filtering).
• Implement new techniques for regression, smoothing, and feature importance analysis.
• Improve robustness of the model for sparse and noisy datasets.
• Integrate and enhance scraping pipelines for structured data ingestion from multiple marketplaces.
• Build and fine-tune an LLM-powered listing helper to extract, classify, and format listing data from user input.
• Support deployment and testing through FastAPI endpoints and Cherry’s internal admin tools.

Tech Stack & Tools

• Python: pandas, numpy, scikit-learn, statsmodels, requests, BeautifulSoup
• FastAPI (for model deployment and API integration)
• Google Cloud Spanner / PostgreSQL
• LLM APIs: OpenAI / Gemini / custom fine-tuned models for NLP
• Version control: GitHub
• Chart generation & analytics via matplotlib / Plotly

Requirements

• Strong experience building and tuning regression models (linear, isotonic, polynomial, or ensemble).
• Excellent understanding of outlier detection (IQR, isolation forest, etc.).
• Proven experience in data preprocessing and feature engineering.
• Familiarity with data scraping and ETL pipelines.
• Experience integrating or fine-tuning LLMs for text-to-structured-data tasks.
• Proficiency with Python ML stack and API integration (FastAPI preferred).
• Analytical mindset and strong documentation habits.

Nice to Have

• Background in automotive or marketplace data modeling.
• Experience with large datasets or commercial valuation systems.
• Familiarity with prompt engineering or LLM evaluation frameworks.

What We Offer

• Fully remote, flexible hours.
• Work with a multidisciplinary product and data team.
• Clear technical documentation and real production data.
• Direct impact on a product used by dealers and transport companies across the US.