Pangram is an AI detection startup that builds models to identify AI-generated and AI-assisted text and images.
About Pangram Labs Pangram Labs’ mission is to protect authenticity by building the world’s best AI detection systems. We publish research on state-of-the-art AI detection techniques and algorithms, and build products to bring our technology into people's daily lives. We are looking for high agency, mission-driven, and passionate builders to join our team. All roles are in person in Brooklyn, New York. About the Role Pangram Labs is hiring AI Research Scientists to join our team. You’ll work directly on our core AI detection model, training deep learning models, running experiments, and contributing innovative ideas to build the most accurate AI detection software in the industry. You will build models that reliably identify AI-generated content and perform research in adjacent areas such as adversarial robustness and interpretability. This is a research role at a startup, which means engineering work is core to the discipline: you'll ship production code, deploy models, build infrastructure, and do real engineering work. Responsibilities: - Train large-scale deep learning models to detect AI-generated content - Experiment with the creation of synthetic datasets at scale - Develop novel and innovative ideas to improve the model’s performance - Share your work through technical publications, whitepapers, and blog posts - Work with engineering and product teams to translate research into new product features and communicate the results to non-technical audiences Requirements: - Ph.D and/or 3-5+ years of experience in AI, Machine Learning, NLP, or related areas - First-author publications in major AI venues such as NeurIPS, ICML, ICLR, etc., or significant contributions to AI products in industry - Deep understanding of generative models and how they work - Experience training neural networks in PyTorch at scale - Familiarity with recent technical developments in LLMs Nice to have - Experience with synthetic data, adversarial robustness, or interpretability - Experience or publications in NLP or computational linguistics - Experience with AI infrastructure / distributed systems About the Team We are a small, tight-knit team of researchers and engineers who care deeply about our work and each other. We believe that a small group of highly talented, empowered, and efficient people can outperform teams many times their size. Everyone on the team has meaningful ownership over the technical and strategic direction of the company. Pangram is primarily an in-person company. We are high-touch, high-energy, and extremely collaborative by nature and choose to spend time together to do our best thinking and innovation.