Data Scientist Engineer Level 2 – Ai/Ml Project Ts/Sci W/Poly

Peraton; Maryland; On-site; USD 90,000 - 120,000; Full time

Job summary

Peraton is seeking a data scientist in Maryland to develop machine learning and statistical algorithms for dataset analysis. This role involves creating data visualizations, collaborating with subject matter experts, and developing models to extract insights.

The position requires a Bachelor's degree in a quantitative field and at least five years of relevant experience, including a strong background in Python and data visualization, as well as familiarity with Large Language Models.

Qualifications

• 5+ years of experience in data analysis and programming with data analysis software.
• PhD can substitute for 4 years of experience.
• Active TS/SCI clearance with polygraph is required.
• Experience optimizing Large Language Models for NLP tasks.

Responsibilities

• Develop machine learning and statistical algorithms to analyze data.
• Generate reports and visualizations providing insights for customers.
• Partner with SMEs for automated analytics translation.
• Evaluate and validate performance of analytics.

Skills

Data analysis with R

Data visualization

Statistical analysis

Machine learning

Large Language Models (LLMs)

GPU-based computing

Python programming

SQL

Education

Bachelor's degree in a quantitative discipline

Tools

Python

R

SAS

MATLAB

Job description

Basic Qualifications

• Bachelor's degree from an accredited college or university in a quantitative discipline (e.g., statistics, mathematics, operations research, engineering or computer science) and five (5) years of experience analyzing datasets and developing analytics as well as five (5) years of experience programming with data analysis software such as R, Python, SAS, or MATLAB.
• An additional four (4) years of experience in software development, cloud development, analyzing datasets, or developing descriptive, predictive, and prescriptive analytics can be substituted for a Bachelor's degree.
• A PhD from an accredited college or university in a quantitative discipline can be substituted for four (4) years of experience.
• Produce data visualizations that provide insight into dataset structure and meaning.
• Work with subject matters experts (SMEs) to identify important information in raw data and develop scripts that extract this information from a variety of data formats (e.g., SQL tables, structured metadata, network logs).
• Incorporate SME input into feature vectors suitable for analytic development and testing.
• Develop AI and machine learning models to address complex problems.
• Develop and optimize Large Language Models (LLM) for various NLP tasks and information retrieval.
• Experience with utilizing GPU‑based computing resources to accelerate model training and deployment.
• Develop and implement statistical, machine learning, and heuristic techniques to create descriptive, predictive, and prescriptive analytics.
• Develop statistical tests to make data‑driven recommendations and decisions.
• Develop experiments to collect data or models to simulate data when required data are unavailable.
• Develop feature vectors for input into machine learning algorithms.
• Identify the most appropriate algorithm for a given dataset and tune input and model parameters.
• Evaluate and validate the performance of analytics using standard techniques and metrics (e.g., cross validation, ROC curves, confusion matrices).
• Oversee the development of individual analytic efforts and guide team in analytic development process.

An Active TS/SCI clearance with polygraph is required.

Additional Desired Qualifications

• AI/ML Integration: Familiarity with Large Language Models (LLMs) and Retrieval‑Augmented Generation (RAG) frameworks, agents, and agentic workflow.

Responsibilities

A data scientist will develop machine learning, data mining, statistical and graph‑based algorithms to analyze and make sense of datasets; prototype or consider several algorithms and decide upon final model based on suitable performance metrics; build models or develop experiments to generate data when training or example datasets are unavailable; generate reports and visualizations that summarize datasets and provide data‑driven insights to customers; partner with subject matter experts to translate manual data analysis into automated analytics; implement prototype algorithms within production frameworks for integration into analyst workflows.