Machine Learning Engineer

BT-436 – Machine Learning Engineer

Location: McLean, VA (fully on-site, no remote option)

Introduction

The Sponsor seeks to develop and adapt Large Language Models (LLM) and Natural Language Processing (NLP) methodologies using artificial intelligence and machine learning (AI/ML) frameworks and programming languages to create new capabilities to improve analytic workflows and address key intelligence questions. The Sponsor needs experienced AI/ML engineering skills to build tools and analytics that use the Sponsor’s structured and unstructured data to yield novel analytic insights, accelerate workflows, enable analysts to surface relevant information from massive data stores. The work will be performed within a team environment.

Work Requirements

• The Candidate shall work closely with the Sponsor’s data scientists and technical team to implement requirements; however, the Sponsor’s GTM will manage the priorities.
• The Candidate shall leverage open source best-in-class models to create new models—through fine tuning, transfer learning, and more— and ensembles that can be re-purposed and reused, covering the following domains:
  • Classification
  • Supervised (CNN, Unet, YOLO, or similar)
  • Unsupervised (DBSCAN, K-Nearest, Adaboost, or similar)
  • Prediction
  • Predictors (LSTM, UConv, or similar)
  • Regressors (Bayesian, linear, or similar)
  • Fusion (sensor level and entity level)
  • Decision making (Task assignments, combinational and parametric optimizations, Reinforcement Learning, Path Planning, or similar)
  • Generative (LLMs, Stackable LLMs, or similar)
  • Machine Learning, AI, Deep Learning (TensorFlow, PyTorch), Text, (Classification, NLP, Topic and Language Modeling, Sentiment Analysis, Information Retrieval), Recommender Systems and Personalization, Threat Detection, Computer Vision, Data Mining, Statistics, or similar.
• The Candidate shall analyze large amounts of raw data, including text data, to determine utility for training or testing AI models.
• The Candidate shall preprocess or clean structured and unstructured Sponsor data, including text data.
• The Candidate shall design and implement advanced ETL code and table configurations for complex data sets.
• The Candidate shall use Structured Query Language (SQL) to develop and organize relevant information.
• The Candidate shall in cooperation with a team, author analytic publications and produce ad-hoc reports to include data visualizations using the Sponsor’s templates, and document methods in Github.
• The Candidate shall implement the Sponsor’s existing coordination process.
• The Candidate shall provide technical education to staff on an ad-hoc basis.
• The Candidate shall provide subject matter expertise in AI/ML to support Sponsor’s initiatives.

Required Skills and Demonstrated Experience

• Demonstrated experience tuning neural networks, such as LLMs, on custom data sets and applying results to specific use cases.
• Demonstrated professional or academic experience developing models and ensembles in the AI/ML space, including selecting the best Python libraries for a given task, choosing appropriate pre-processing actions, performing analysis, and assessing model performance.
• Demonstrated professional or academic experience using Python or R.
• Demonstrated professional or academic experience with deep learning frameworks such as PyTorch, Tensorflow, or Keras.
• Demonstrated professional or academic experience and proficiency with SQL to include using common table expressions, set operations, aggregated functions and nested subqueries.
• Demonstrated professional or academic experience with version control systems such as Github and Jenkins.
• Demonstrated experience leveraging GPUs for accelerated computing.

Highly Desired Skills and Demonstrated Experience

• Demonstrated professional or academic experience with the HuggingFace Transformers library and hub.
• Demonstrated experience with cloud computing development and architecture.
• Demonstrated experience with front-end web development frameworks such as Flask.
• Demonstrated experience creating machine learning models that conduct text classification and topic modeling in Python using standard machine learning (SciKit Learn) or deep learning models.
• Demonstrated experience developing applications for semantic search.
• Demonstrated professional or academic experience and proficiency with Tableau to produce visualizations and dashboards.
• Demonstrated academic or professional experience communicating methodological choices and model results.
• Demonstrated experience with verification and validation test benches.
• Demonstrated experience with Explainable AI (XAI) techniques.
• Demonstrated experience with ONNX (Open Neural Net Exchange).