Data Scientist

Pacific Air Industries; Los Angeles (CA); Hybrid; USD 120,000 - 190,000; Full time

Job summary

Pacific Air Industries, a leading aviation aftermarket business, seeks a Data Scientist to develop machine learning, pricing, valuation, and predictive analytics that influence revenue, margins, purchasing, and sales.

You will work with Snowflake data across more than two million part numbers, building models for pricing, package valuation, quote prioritization, and inventory optimization. This role blends data analytics, ML, SQL, and business problem solving to deliver measurable results.

Qualifications

• Strong grounding in statistics and predictive modeling.
• Experience deploying models in production that drive business value.
• Proficient with SQL and Python, and comfortable with Snowflake data.

Responsibilities

• Develop machine learning, pricing, valuation, and predictive analytics solutions to impact revenue and margins.
• Build and improve models using Snowflake data across 2M part numbers.
• Create self-service tools for Sales and Operations to access accurate data for RFQs and quotes.
• Evaluate model performance against actual results and iterate accordingly.
• Collaborate across Data Science, Sales, and Operations to translate data into actionable decisions.

Skills

Statistical methods

Machine learning

Communication

Problem solving

Team collaboration

Education

Bachelor’s degree in a technical field

Tools

SQL

Python

Snowflake

dbt

Job description

Location: Hybrid (3 days onsite) 9650 De Soto Avenue, Chatsworth, CA 91311

About the Company

Pacific Air Industries (Pac-Air) and Air-Cert are two aviation aftermarket businesses under common ownership. Pacific Air buys surplus expendable aircraft parts from airlines and MROs and redistributes them, pricing across a part master of more than two million part numbers with roughly 400,000 of them in stock at any time. Air-Cert is an FAA and EASA certificated repair station that repairs aircraft components for commercial airlines, freight carriers and aerospace OEMs. Pacific Air has been in business since 1959 and Air-Cert since 1962. Nearly every decision either company makes is a judgment about what a part is worth, what to pay for it, and when to buy it, at a scale that needs excellent data management.

Today we are profitable and growing, and the models this role owns sit under the largest capital decisions we make. Our Snowflake environment holds roughly 24 million rows across 134 objects, keyed to that part master. This is our first data science hire, so you would build the function instead of inheriting it, working across both businesses. The work is turning data into insights, insights into decisions, and decisions into profit.

Overview

We are seeking a Data Scientist to develop machine learning, pricing, valuation, and predictive analytics solutions that directly impact revenue, margins, purchasing, and sales performance. Working with Snowflake data across more than two million part numbers, this role will build and improve models for part and repair pricing, package valuation, quote prioritization, purchasing recommendations, inventory optimization, and demand forecasting. The role will evaluate model performance against actual business results and apply machine learning to improve existing rules-based processes.

This is a highly business-focused data science role that combines data analytics, machine learning, statistical analysis, SQL, and business problem-solving. The Data Scientist will also develop self-service tools that give Sales and Operations direct access to accurate, actionable data for RFQs, customer quotes, purchasing decisions, and customer meetings—turning complex data into practical solutions that drive measurable business results.

Key Accountabilities

• The part-level data model. The pricing, inventory, and transaction models in Snowflake, and the third-party market data they draw on, across both companies and more than two million part numbers. That covers their logic, documentation, and testing, and making sure sales figures tie to our monthly reporting.
• Pricing. Every Pacific Air part price and every Air-Cert repair price: how it is calculated, how it updates part by part as new quote and market data arrives, and how it performs against actual sales and quote outcomes. That includes moving from rules to trained models wherever a model beats the rule, on both the buy side and the sell side, and using price to prioritize quote responses.
• Package valuation. The models behind every package bid, expendable and rotable, and the ongoing check of what each package was expected to earn against what it actually earned.
• Buying recommendations. The recommendations on which parts justify a speculative buy and when and how much to reorder, based on inquiry and quote activity, typical PO size, lead time, demand history, and margin, and how those buys actually perform. Also the Air-Cert parts purchase forecast built from daily inbound work orders.
• Sales access to the models. The self-serve query tools being built for the sales team, recurring jobs such as pricing a customer's RFQ file, and data prepared for sales ahead of customer visits. The measure is whether sales gets correct answers without asking someone to pull the data.

First Year Priorities

• Take over the pricing, inventory, and transaction models already rebuilt in Snowflake, and run them independently.
• Rebuild the expendable package model against realized results and build the first rotable valuation model.
• Put a trained pricing model into the live quoting flow, including quote response prioritization.
• Build & automate the speculative buy list and recurring reorder recommendations.
• Build RFQ pricing, so sales can upload a customer's RFQ and get priced lines back.
• Extend pricing and parts forecasting to Air-Cert.

We expect the first two or three to take most of the year.

Qualifications

• Bachelor’s degree in a technical field, including but not limited to: Computer Science, Database Management, Mathematics, or other closely related areas
• 3+ years of professional experience building predictive models, including at least one pricing, forecasting, or inventory model that reached production and that a business relied on.
• Knowledge and practical application of statistical methods.
• Demonstrated practical experience applying machine learning to real-world business problems, including developing, testing, validating, and deploying predictive models using historical and continuously changing datasets preferred.
• Experience with SQL and Python strongly preferred.
• Knowledge of Snowflake and data build tool preferred.

Additional Skills and Attributes

• Strong verbal and written communication, presentation, and interpersonal skills.
• Ability to work collaboratively or independently; demonstrable bias for organization and time management with proven capacity for completing work on schedule
• Capable of working on several projects simultaneously, demonstrating flexibility in accommodating shifts in prioritization as needed
• Ability to work in a stationary position 70% of the time, consistently operating a computer and office productivity equipment; occasionally moves around an office environment; ability to communicate and accurately exchange information both visually and audibly