Apply on Employer Site

Steampunk, Inc. · 1 month ago

Data Scientist- Generative AI

US-VA-McLean

Full-time

Onsite

Entry, Mid Level

$125K/yr - $160K/yr

2+ years exp

Steampunk, Inc. is a Change Agent in the Federal contracting industry, seeking a seasoned Data Scientist to work with their team on Generative AI technology for Federal use cases. The role involves designing and training advanced machine learning models, collaborating with engineers, and staying updated with AI trends to implement innovative solutions.

ConsultingInformation Technology

Growth Opportunities

No H1B

Security Clearance Required

U.S. Citizen Only

Responsibilities

Design and train advanced machine learning models, especially generative models like GANs, VAEs, or transformer-based models

Work closely with ML Engineers and Software Developers to transition models from a research and development stage to production

Stay updated with the latest research and trends in AI to implement cutting-edge solutions

Evaluate the performance of foundational models (e.g., GPT, LLaMA, Claude) on domain-specific tasks and fine-tune them using supervised, reinforcement, or instruction tuning methods to align outputs with user needs and business goals

Design and optimize prompts, few-shot examples, and system instructions to improve LLM behavior in constrained environments (e.g., RAG pipelines, multi-agent workflows, decision support systems)

Identify, clean, label, and synthesize high-quality datasets for model training, fine-tuning, or retrieval-augmented generation (RAG)

Design experiments to evaluate generative model behavior (e.g., hallucination rates, factuality, coherence, safety), define appropriate benchmarks, and use metrics like BLEU, ROUGE, perplexity, and human evaluations

Ability to leverage and integrate various data management tools at scale – cloud experience preferred

Support an Agile software development lifecycle

You will contribute to the growth of our AI & Data Exploitation Practice!

Ability to hold a position of public trust with the US government

Qualification

Generative AIMachine LearningPythonNatural Language ProcessingCloud AnalyticsData ManipulationAI FrameworksModel EvaluationData VisualizationSoft Skills

Required

Ability to hold a position of public trust with the US government

Bachelor's degree in computer science, data science/statistics, information systems, engineering, business, or a scientific or technical discipline

2-4 years industry experience developing ML/AI solutions and a passion for solving complex problems

Must have hands-on experience in building and deploying generative models, with a portfolio of relevant projects

Must be proficient in Python, with strong coding practices for scalability and reproducibility

Demonstrated experience in building, training, and deploying generative models like Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), or autoregressive models like Transformer-based architectures

Demonstrated experience manipulating structured and unstructured data for analysis

Demonstrated experience using AI frameworks like TensorFlow, PyTorch, Keras, or JAX

Ability to implement and modify complex neural network architectures

Skilled in using data manipulation and analysis libraries such as Pandas, NumPy, SciPy, and Scikit-learn

Experience with model training, fine-tuning, and evaluation using frameworks like PyTorch or TensorFlow

Deep knowledge of natural language processing techniques, including tokenization, embeddings, attention mechanisms, and prompt engineering

Hands-on experience with large language models (e.g., GPT, LLaMA, Claude, Mistral) and associated libraries (e.g., Hugging Face Transformers)

Familiar with retrieval-augmented generation (RAG), data labeling, synthetic data generation, and data governance best practices

Proficiency in Python and ML tooling (e.g., Jupyter, Git, Docker, APIs). Able to work in cross-functional teams to deliver AI capabilities into production environments, and write modular/reusable code

Experience in Cloud analytics (AWS, Azure, or Google Cloud Platform - GCP) with tools such as AWS SageMaker, AWS Bedrock, Azure OpenAI, etc

Experience with DevSecOps, as it applies to data science and MLOps

Preferred

Cloud experience preferred

Data visualization skills in Tableau, Power BI, D3, ArcGIS, or similar are a plus

Experience with Elasticsearch, AWS Kendra, Azure Cognitive Search, or similar tool is a plus

Benefits

Steampunk’s total compensation package for employees.

Company

Steampunk, Inc.

Glassdoor4.4

Steampunk is anchored by a startup culture with a customer-centered delivery approach, we put our Federal government clients in the center of everything we design, develop, and deliver to drive high-quality mission impacts and user experiences at speed.

Founded in 2019

Washington, District of Columbia, USA

201-500 employees