Instacart is looking to unify search, discovery, merchandising, and recommendations under a single, reasoning-rich AI platform by consolidating over 80 task-specific models into a small set of large, general-purpose LLMs and multi-task models.
Requirements
- Have 5+ years of industry experience using machine learning to solve real-world problems with large datasets with 3+ years in a technical leadership role
- Proven track record designing and deploying sophisticated ML/AI systems in production environments that drive measurable business impact through improved recommendations, search relevance, and user engagement metrics.
- Have strong programming skills in Python and fluency in data manipulation (SQL, Pandas) and Machine Learning (scikit-learn, XGBoost, Keras/Tensorflow) tools
- Have strong analytical skills and problem-solving ability
- Extensive expertise with modern deep learning frameworks (PyTorch, TensorFlow, JAX) and advanced LLM architectures including transformer models, attention mechanisms, and multimodal AI systems.
- Demonstrated experience implementing and fine-tuning large language models, including prompt engineering, embedding techniques, and efficient inference optimization for production environments.
- Strong foundation in AI fundamentals including neural network architectures, generative models, and foundation model adaptation methodologies like PEFT, LoRA, and RLHF.
Responsibilities
- Architect and scale foundational LLM systems that unify query understanding, personalization, ranking, ads, and merchandising—replacing dozens of siloed models with a single, adaptive backbone.
- Develop agentic workflows that proactively support shoppers by generating personalized cart starters, thematic bundles, and dynamic storefronts—driven by LLMs reasoning over historical behavior, preferences, and context.
- Fine-tune large language models using SFT, DPO, and GRPO on rich behavioral feedback signals to align system outputs with evolving customer needs and business objectives.
- Design intelligent, end-to-end discovery pipelines that handle everything from long-tail semantic retrieval to real-time multi-objective ranking and merchandising optimization.
- Collaborate cross-functionally with product, design, infra, and ads teams to translate high-level discovery goals into scalable LLM-powered systems that drive measurable impact.
- Coach and mentor a team of ML engineers, fostering their technical and professional growth
Other
- Are a strong communicator who can collaborate with diverse stakeholders across all levels
- Self-motivated innovator with a strong sense of ownership who can navigate the rapidly evolving AI landscape, evaluate emerging techniques, and implement novel approaches to solve complex business challenges.
- Passion for applying cutting-edge AI research to real-world applications and a keen understanding of the practical considerations in developing responsible, efficient AI systems at scale.
