Senior Data Scientist / Boston MA( Hybrid),6+ months contract

Job Posting

Dynamic work schedule: This is 5 days on site a month in the same week then the remainder of the month is working from home. If your candidate is not open to this please tell them they will not be considered. Client determines the week they are working onsite, not the candidate.

Skills:

• Focus on NLP/NLG, LLMs, ASR, FLAN/T5, Llama, BERT, BART, LSTMs, BLOOM, Pythia, DOLLY, ASR, etc.
• Someone who is really good with Python.
• This is for a senior level candidates who can work independently as Data Scientists.

Designing and building critical AI capabilities and services that will benefit institutional and retail clients for years to come. We work closely with all businesses, helping stakeholders navigate through the complexities and nuances of artificial intelligence and advanced analytics to deliver high value, scalable solutions that drive customer and business value. You will be joining an organization that loves whitepapers, code, enjoys data, and thrives in a fast-paced environment solving the toughest AI problems in the financial industry. The AI team you'll be joining contributes to the vitality and growth of the organization by researching and building complex, scalable AI algorithms, models, frameworks, platforms and technologies to significantly improve customer experience and drive business results. Our team of career data scientists, mathematicians and engineers takes a strategic and rigorous approach to ensure Fidelity continues to push the boundaries of AI in the financial industry.

The Expertise You Bring: A PhD or master's degree in computer science, computational statistics, mathematics, physics, engineering, optimization or related applied mathematics and technical disciplines as they pertain to data science/machine learning (Client). Academic publications are plus, but not required. You move fluidly between model development and software engineering – including model evaluation, tuning/performance, operationalization, Client demos and visualizations, as well as scaling algorithms. This is neither a pure research role nor an Client ops/engineering position, but you're comfortable building your own pipelines and partnering with technology to help streamline model deployment. Experience developing a range of models over the course of your professional career. Mastery of foundational/classical machine learning is assumed (regression, random forest, gradient boosting, ensembles, neural networks, feature selection/reduction, parameter tuning, etc.), however, expertise in one of the following Client/CS solution domains will be given preference: NLP/NLG, LLMs, ASR, embedding services/stores, i.e., transformer-based capabilities / architectures and variants (any modality, e.g., text, image, audio); trending open-source model genres, e.g., GPT, FLAN/T5, Llama, BERT, BART, LSTMs, BLOOM, Pythia, DOLLY, ASR, etc. Production recommender systems (Bayesian models, collaborative systems, MABs, and reinforcement learning). Experience writing production-level code using cloud-based environments, pipelines and frameworks, e.g., AWS SageMaker/endpoints, EKS/Kubernetes, Kinesis, MS Azure, GCP and PEP 8 coding standards; prototyping and deploying Client solutions, using experimentation design and GLMs. Experience on projects involving a variety of massive data sources – from relational data to unstructured data such as clickstream, text, images and audio. Expert programming, framework, and technology skills in several of the following: Python, PyTorch, TensorFlow (optional), JAX (optional), Streamlit, Gradio, Spark/PySpark (optional), SQL, Containerization/Docker, Hugging Face, git/git stash. C/C++/C# experience is a plus, even if you have not used these skills recently.

Good to Have: Experience with full stack software development would be desirable. Real-time event detection and scoring a plus. CUDA/GPU optimization. A strategic mindset. Excellent communication/presentation skills and comfortable working with teams.

The Value You Deliver: You will be a trusted AI partner helping business, product, and engineering stakeholders translate business requirements into real-world next-generation AI-driven solutions. The ideal candidate combines expert open-source AI/Client modeling with hands-on technical experience to build algorithms, models and, occasionally, applications that surface insights and deliver value (revenue and/or cost saving benefits, customer experiences/journeys, and increased efficiencies through automation/optimization).