Distinguished Data Scientist

Fredericton, NB (Eastern Canada)

As a distinguished data scientist at TrojAI you will be a key member of the data science leadership team helping guide the productization of AI security protections for both discriminative and generative AI models within the TrojAI platform. If you are looking to technically engage with the largest enterprises in the world, helping them securely adopt AI, this role may be for you.

About the Role

The distinguished data scientist will work closely with engineering and product teams as a hands-on technical expert in the AI security space. In addition to being an individual contributor, this role will help scope out the predictable development and delivery of ongoing AI security protections, providing mentorship across the data science team.

In this role you will:

• Help lead the data science team fostering innovation and excellence in AI security engineering translating business needs into short, medium and long-term roadmaps
• Develop technical assets and IP across automated pentesting and real-time monitoring of both discriminative and generative AI technologies
• Engage with product owners to influence product direction and facilitate coverage across security frameworks and standards (e.g., NIST AI RMF, OWASP TOP 10s, MITRE ATLAS, CLOUD SECURITY ALLIANCE, etc.)
• Track and analyze emerging threats to AI systems, focusing on AI/ML/GenAI models, applications and environments
• Develop and implement detection and mitigation strategies for identified threats, including prototyping new approaches
• Collaborate with product teams to integrate and operationalize quantifiable solutions
• Contribute to a strong culture of innovation, delivery, impact, and continuous improvement
• Create field enablement materials to educate technical field personnel on integrating AI protections into customer workflows
• Develop white papers, blogs, reference implementations, and presentations to facilitate customer and partner adoption of AI protections
• Represent data science as thought leaders in the AI security space at industry events and conferences

The skills required for this role:

• Deep understanding of AI/ML models, including their strengths, limitations, risks and vulnerabilities
• Experience with AI/ML security risks such as data poisoning, privacy attacks, adversarial inputs, denial of service attacks, etc.
• Excellent written and verbal communication skills, paired w. strong analytical and problem-solving skills
• Strong programming skills in generic programming languages such as Python or Golang and machine learning frameworks such as TensorFlow or PyTorch
• Ability to quickly learn new technologies and concepts and to understand a wide variety of technical challenges to be solved
• Hands-on experience with AI security best practices, including vulnerability assessments, red teaming, and cybersecurity controls
• 7+ years of experience as a Data Scientist or Machine Learning Solutions Architect, preferably with a focus on AI/ML security
• 5+ years of proficiency in Python for data analysis, model training, evaluation, deployment, and optimization
• 1+ year of experience with technologies related to large language models (LLM architectures, responsible GenAI, model evaluation, and customization techniques)

Qualifications:

• Experience with open source frameworks for building applications powered by large language models like LangChain and LlamaIndex
• Knowledge of prompt engineering, state-of-the art model architectures as well as necessary understanding of underlying maths/statistics
• Familiarity with AWS, Azure & Google services and the cloud computing landscape
• Knowledge of model deployment technologies such as TorchServe, ONNX, Triton, vLLM and orchestration using docker and Kubernetes is a plus
• Familiarity with big data technologies such as Hadoop, Spark, and MongoDB
• Demonstrated ability to think strategically about business, product, and technical challenges in an enterprise environment
• Track record of thought leadership and innovation around AI/ML/GenAI, ideally with experience in patenting inventions
• Experience developing proof-of-concept exploits for new or theoretical attacks
• Experience scaling products at hyper-growth startups
• Graduate degree in a quantitative field such as statistics, mathematics, data science, engineering, or computer science

About TrojAI

TrojAI enables the secure use of artificial intelligence for the largest enterprises in the world. The TrojAI platform assesses risks of AI models across traditional AI/ML and GenAI and protects applications from active threats. An original AI security company, TrojAI was founded in 2019 by data scientists and cybersecurity professionals with the mission of enabling the secure use of AI.