Harvard Business School Foundry is seeking to build, fund, and launch ventures through an AI-driven digital learning platform. The initiative aims to develop tools and experiences that empower founders to grow their ventures and confidence, combining the energy of a startup with the mission and reach of Harvard.
Requirements
- Minimum of five years’ software development experience with Python and SQL.
- Minimum of three years’ experience building pipelines to deploy NLP and deep learning models into production in a cloud environment.
- Minimum three years’ experience using PyTorch, Tensorflow, or MXNet, along with optimizing code for GPU clusters.
- Experience building advanced workflows such as retrieval augmented generation, model chaining, dynamic prompting, PEFT/SFT, etc. using Langchain and similar tools.
- Experience establishing model guardrails and developing bias detection and mitigation techniques for AI applications using tools such as NeMo.
- Experience with various embedding models and setting up and tuning vector databases to improve performance of semantic search and retrieval systems.
- Understand the underlying fundamentals such as Transformers, Self-Attention mechanisms that form the theoretical foundation of LLMs.
Responsibilities
- Architect, build, maintain, and improve new and existing suite of GenAI applications and their underlying systems.
- Automate machine learning pipelines, monitor performance and costs, and optimize models by using techniques such as LoRA/QLoRA.
- Establish reusable frameworks to streamline model building, deployment and monitoring. Incorporate comprehensive monitoring, logging, tracing, and alerting mechanisms.
- Build guardrails, compliance rules and oversight workflows into the GenAI application platform, such as establishing approval chains for model updates and staged rollout for production releases.
- Develop templates, guides and sandbox environments for easy onboarding of new contributors and experimentation with new techniques.
- Ensure development of user-facing applications in the GenAI application platform is easy and safe by enforcing rigorous validation testing before publishing user-generated models and implement a clear peer review process of applications.
- Work closely with data scientists and analysts to create and deploy new product features online and in mobile apps.
Other
- Minimum of seven years’ post-secondary education or relevant work experience.
- Bachelor's degree in mathematics, physics, computer science, engineering, statistics, or an equivalent technical discipline desired.
- Harvard University is unable to provide visa sponsorship for this position.
- This is a hybrid position which we consider to be a combination of remote and onsite work at our Boston, MA based campus. HBS expects all staff to be onsite a minimum of 3 days per week.
- A cover letter is required to be considered for this opportunity.
