The internship is based at the company's headquarters in Mountain View, California.
The company runs a large professional network that helps members find opportunities, build skills and make connections.
Recommender teams choose which posts, videos and connections each user sees, aiming to keep content fresh so people come back.
You will work with very large volumes of semi-structured text, graph and user activity data to build scalable, intuitive recommender systems for the feed, jobs and learning products.
You will learn alongside experienced engineers and researchers to design recommendation methods that make personalization more relevant for millions of users.
Responsibilities
Run research and development on advanced recommender methods, such as collaborative filtering, matrix factorization, deep learning and reinforcement learning.
Build scalable algorithms that personalize the platform while balancing relevance, diversity and fairness.
Work with engineering and product teams to put your solutions into the platform, reaching users worldwide.
Use large datasets to train and assess recommender models, refining them until performance is strong.
Work with mentors, business specialists and technologists, doing independent research and helping deliver clear solutions to products and services.
Requirements
Basic: You are enrolled in a PhD in computer science, statistics, mathematics, electrical engineering, machine learning or a closely related technical field, and you will return to it after the internship.
Basic: Familiarity with recommender systems, machine learning or adjacent areas.
Basic: Skill in Python and in ML libraries such as TensorFlow or PyTorch.
Basic: Understanding of core recommender techniques, including collaborative filtering, content-based methods, hybrid models and deep learning approaches.
Basic: Knowledge of recommendation quality metrics such as precision, recall, AUC and diversity.
Preferred: Experience with languages used in AI and large-scale systems, such as Python, Java, C++ and Go.
Preferred: Familiarity with data processing frameworks like Apache Spark, Ray, Flink or Databricks, plus distributed computing concepts such as MapReduce and cloud-native pipelines.
Preferred: Experience shipping recommender or large-scale ML models to production, using methods like embeddings, graph neural networks or multi-task learning.
Preferred: Knowledge of reinforcement learning and reinforcement learning with human feedback applied to recommendation or personalization.
Preferred: Familiarity with LLM-based or hybrid retrieval and ranking systems.
Preferred: Skill with modern ML frameworks such as TensorFlow, PyTorch, JAX, Hugging Face Transformers, Scikit-Learn, NumPy and Pandas.
Preferred: Knowledge of cloud ML platforms (AWS SageMaker, GCP Vertex AI, Azure ML) and MLOps tools (MLflow, Kubeflow, Weights & Biases).
Preferred: Research contributions or publications at venues such as NeurIPS, ICML, ICLR or KDD.
Preferred: Clear communication and teamwork, with the ability to turn complex technical ideas into business impact.
Suggested: Background in machine learning and deep learning research.
Suggested: Experience with large datasets and data mining.
Suggested: Strategic thinking and problem-solving ability.
Conditions
Choose one of two 12 week sessions: May 24, 2027 to August 13, 2027, or June 14, 2027 to September 3, 2027.
You must be enrolled in a PhD program with an expected graduation date of December 2027 or later.
The work is hybrid: you split time between home and a company office on select days, set by the team's business needs.
Base pay is $62 to $75 per hour, depending on skill set, qualifications, certifications and work location, and may differ in other locations due to labor costs.
The company is a federal contractor and follows pay transparency and non-discrimination rules.
Benefits
Total pay may also include an annual performance bonus, stock, benefits and other applicable incentive plans.
Benefit details are available on the company's careers benefits page.