Perplexity

Perplexity is seeking an experienced Machine Learning Research Engineer to help build the next generation of advanced search technologies, with a focus on retrieval and ranking. Responsibilities - Relentlessly push search quality forward — through models, data, tools, or any other leverage available - Architect and build core components of the search platform and model stack - Design, train, and optimize large-scale deep learning models using frameworks like PyTorch, leveraging distributed training (e.g., PyTorch Distributed, DeepSpeed, FSDP) and hardware acceleration, with a focus on retrieval and ranking models - Conduct advanced research in representation learning, including contrastive learning, multilingual, and multimodal modeling for search and retrieval - Deploy models — from boosting algorithms to LLMs — in a scalable and performant way - Build and optimize RAG pipelines for grounding and answer generation - Collaborate with Data, AI, Infrastructure, and Product teams to ensure fast and high-quality delivery Qualifications - Deep understanding of search and retrieval systems, including quality evaluation principles and metrics - Proven track record with large-scale search or recommender systems - Strong proficiency with PyTorch, including experience in distributed training techniques and performance optimization for large models - Expertise in representation learning, including contrastive learning and embedding space alignment for multilingual and multimodal applications - Strong publication record in AI/ML conferences or workshops (e.g., NeurIPS, ICML, ICLR, ACL, CVPR, SIGIR) - Self-driven, with a strong sense of ownership and execution - Minimum of 3 years (preferably 5+) working on search, recommender systems, or closely related research areas

Perplexity is seeking top-tier AI Research Scientists and Engineers to advance our AI products and capabilities. We're building the future of AI-powered search and agent experiences through our Sonar models, Deep Research Agent, Comet Agent, and Search products. Join us in creating SOTA experiences that handle hundreds of millions of queries and continue to scale rapidly. Team Structure Depending on your interests and expertise, you'll work on one of three specialized teams: 1. Core Research Team (Horizontal) Focus on generating and improving base models that power all our products. This team works on foundational model capabilities, post-training techniques, building RL infra and infrastructure that benefits the entire organization. 2. Agent Products Team (Vertical) Concentrate on fine-tuning and optimizing models for our Deep Research Agent and Labs/Canvas products. This team bridges research and product, ensuring our agent capabilities deliver exceptional user experiences. 3. Comet Agent Team (Vertical) Dedicated to developing and enhancing our Comet Agent product. This specialized team focuses on the unique requirements and optimizations needed for Comet's specific use cases. Responsibilities Research & Development - Post-train SOTA LLMs using the latest supervised and reinforcement learning techniques (SFT/DPO/GRPO) - Leverage our rich query/answer dataset to scale model performance across Sonar, Deep Research, Comet, and Search products - Stay current with the latest LLM research, especially in model training, optimization, and personalization techniques - Implement preference optimization and personalization capabilities to enhance user experience - Invent in-house improvements and optimizations to enhance SOTA models - Turn research ideas into algorithms and run experiments to launch new models Infrastructure & Implementation - Own full-stack data, training, and evaluation pipelines required for model development - Build robust and effective training frameworks (on top of Megatron/PyTorch) for post-training LLMs - Implement necessary infrastructure and components to support cutting-edge model training at scale - Integrate models seamlessly into our product ecosystem Collaboration - Work closely with engineering teams to integrate models into Perplexity's product suite - Collaborate across teams to ensure cohesive AI experiences throughout our platform - Partner with product teams to understand user needs and translate them into model improvements Qualifications Required - Proven experience with large-scale LLMs and Deep Learning systems - Strong programming skills in Python/PyTorch; versatility is a plus - Experience with post-training techniques and reinforcement learning - Self-starter with a willingness to take ownership of tasks - Passion for tackling challenging problems - Minimum 2-6 years of experience on relevant projects (depending on seniority level) Nice-to-have - PhD in Machine Learning, AI, Systems, or related areas - Experience in post-training LLMs with SFT/DPO/GRPO - C++/CUDA programming skills - Experience building LLM training frameworks - Academic publications and research impact - Experience with agent systems and multi-step reasoning - Background in personalization and preference learning

Perplexity is seeking a Research Engineering Manager to lead the team of all-star AI researchers and engineers responsible for developing the models that drive our products. Our team has developed some of the most advanced models for agentic research, query understanding, and other domains that require accuracy and depth. As we expand our userbase and portfolio of product surfaces, our in-house models are increasingly critical to providing a premium, high-taste experience for the world’s most sophisticated users. You will dive into our rich datasets of conversational and agentic queries, leveraging cutting‑edge training techniques to scale AI model performance. Through hands-on technical and organizational leadership, you will empower your team to develop SotA models for the use cases that matter most to our business and our users. RESPONSIBILITIES - Lead a team of researchers and engineers focused on training SotA models for Perplexity-relevant use cases, leveraging the latest supervised and reinforcement learning techniques. - Drive research and engineering efforts to develop production models through advanced model training and alignment techniques, including RL, SFT, and other approaches. - Become deeply familiar with the team’s technical stack, leading from the front through hands-on technical contributions. - Own the data, training, and eval pipelines required to train and continuously improve LLM models. - Design and iterate on model training and finetuning algorithms (e.g., preference‑based methods, reinforcement learning from human or AI feedback) through an approach that balances scientific rigor and iteration velocity. - Design evaluations and improve the production model training pipeline to reliably deliver models that lie on the Pareto frontier of speed and quality. - Work closely with engineering teams to integrate in-house models into our product and rapidly iterate based on real‑world usage. - Manage day‑to‑day execution, project planning, and prioritization for the model training team to hit ambitious quality and performance goals. QUALIFICATIONS - Proven experience with large-scale LLMs and Deep Learning systems. - Strong Python and PyTorch skills; versatility across languages and frameworks is a plus. - Experience leading or managing research or engineering teams working on large-scale AI model development, including driving complex projects from idea to production. - Self‑starter with a willingness to take ownership of tasks and navigate ambiguity in a fast‑moving environment. - Passion for tackling challenging problems in AI model quality, speed, safety, and reliability. - 10+ years of technical experience, with at least 2 of those years as a manager and at least 4 of those years working on large-scale AI model development. NICE-TO-HAVE - PhD in Machine Learning or related areas. - Experience training very large Transformer-based models with techniques such as SFT, DPO, GRPO, RLHF‑style methods, or related preference‑based optimization approaches. - Prior experience designing evaluations and production training pipelines for large‑scale models in a high‑growth environment.

Perplexity is seeking an experienced Machine Learning Research Engineer to help build the next generation of advanced search technologies, with a focus on retrieval and ranking. Responsibilities - Relentlessly push search quality forward — through models, data, tools, or any other leverage available - Architect and build core components of the search platform and model stack - Design, train, and optimize large-scale deep learning models using frameworks like PyTorch, leveraging distributed training (e.g., PyTorch Distributed, DeepSpeed, FSDP) and hardware acceleration, with a focus on retrieval and ranking models - Conduct advanced research in representation learning, including contrastive learning, multilingual, and multimodal modeling for search and retrieval - Deploy models — from boosting algorithms to LLMs — in a scalable and performant way - Build and optimize RAG pipelines for grounding and answer generation - Collaborate with Data, AI, Infrastructure, and Product teams to ensure fast and high-quality delivery Qualifications - Deep understanding of search and retrieval systems, including quality evaluation principles and metrics - Proven track record with large-scale search or recommender systems - Strong proficiency with PyTorch, including experience in distributed training techniques and performance optimization for large models - Expertise in representation learning, including contrastive learning and embedding space alignment for multilingual and multimodal applications - Strong publication record in AI/ML conferences or workshops (e.g., NeurIPS, ICML, ICLR, ACL, CVPR, SIGIR) - Self-driven, with a strong sense of ownership and execution - Minimum of 3 years (preferably 5+) working on search, recommender systems, or closely related research areas

Perplexity is looking for a Search Machine Learning Engineer Intern to help build the next generation of advanced search technologies, with a focus on retrieval and ranking. You will work closely with experienced engineers to improve search quality, experiment with new models, and ship features that directly impact how users search and discover information. Internship program: 12 - 24 weeks, full-time, in-person in the Belgrade office. Responsibilities: - Contribute to experiments that improve search quality through better models, data usage, and evaluation tools, under the guidance of senior engineers. - Design and implement components of the search platform and model stack, including retrieval, ranking, and classification models. - Train evaluating models (including LLM-based approaches) for retrieval, ranking, and classification tasks. - Support deployment and monitoring of search and ranking models in a scalable and performant way. - Help build and iterate on RAG pipelines for grounding and answer generation. - Collaborate with Data, AI, Infrastructure and Product teams to deliver improvements quickly and learn best practices in production ML. Qualifications: - Strong foundation in machine learning and statistics, with coursework or projects related to information retrieval, ranking, or recommender systems. - Experience with Python and common ML frameworks (e.g. PyTorch, TensorFlow, JAX) through academic, open source, or personal projects. - Familiarity with evaluating model quality using offline metrics and/or A/B testing is a plus, but not required. - Previous experience (internships, research, or significant projects) working on search, recommendation, or NLP is a plus, but not required. - Self-driven and curious, with a strong sense of ownership, willingness to learn, and comfort working in a fast-paced environment - Experience with Rust will be a plus

Internship Program Berlin Internship program: 12 - 24 weeks, full-time, in-person in the Berlin office. Responsibilities - Relentlessly push search quality forward — through models, data, tools, or any other leverage available. - Train, and optimize large-scale deep learning models using frameworks like PyTorch, leveraging distributed training (e.g., PyTorch Distributed, DeepSpeed, FSDP) and hardware acceleration, with a focus on retrieval and ranking models. - Conduct research in representation learning, including contrastive learning, multilingual, evaluation, and multimodal modeling for search and retrieval. - Build and optimize RAG pipelines for grounding and answer generation. Qualifications - Understanding of search and retrieval systems, including quality evaluation principles and metrics. - Strong proficiency with PyTorch, including experience in distributed training techniques and performance optimization for large models. - Interested in representation learning, including contrastive learning, dense & sparse vector representations, representation fusion, cross-lingual representation alignment, training data optimization and robust evaluation. - Publication record in AI/ML conferences or workshops (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP, SIGIR).

Perplexity is seeking an exceptional AI Research Tech Lead to drive our research strategy and lead the development of our in-house Online LLMs, the Sonar models. In this leadership role, you will set the macro research direction across different modalities, mentor a team of researchers, and take advantage of our rich query/answer dataset to continue scaling our Sonar model performance and deliver the SOTA Online LLM experience to our users. RESPONSIBILITIES RESEARCH LEADERSHIP & STRATEGY - Define and execute the macro research direction across multiple modalities, including post-training LLMs for agent trajectories and future mid-training initiatives - Lead strategic research planning and roadmap development to advance Sonar model capabilities - Drive innovation in supervised and reinforcement learning techniques for query answering - Collaborate with leadership to align research priorities with product and business objectives TEAM DEVELOPMENT & MENTORSHIP - Coach and mentor a team of AI research scientists and engineers, fostering their technical and professional growth - Establish the long-term macro research direction across the team, including our direction across different modalities - Lead hiring and onboarding of new research talent - Create a collaborative environment that encourages knowledge sharing and innovation TECHNICAL EXCELLENCE - Post-train SOTA LLMs on query answering using cutting-edge supervised and reinforcement learning techniques - Own and optimize the full stack data, training, and evaluation pipelines required for LLM post-training - Deliver Sonar models that provide SOTA query answering performance - Drive research into agent trajectories and multi-modal capabilities - Lead the technical roadmap for eventual mid-training investments CROSS-FUNCTIONAL COLLABORATION - Work closely with engineering teams to integrate Sonar models into our product - Partner with product teams to understand user needs and translate them into research priorities - Collaborate with data teams to leverage our unique query/answer dataset effectively - Communicate research progress and findings to stakeholders across the organization QUALIFICATIONS REQUIRED - Minimum of 5 years of experience working on relevant AI/ML projects with 3+ years in a technical leadership role - Proven track record of leading and mentoring technical and research teams - A Computer Science graduate degree at a premier academic intitution - Deep expertise with large-scale LLMs and Deep Learning systems - Strong programming skills with versatility across multiple languages and frameworks - Demonstrated ability to set technical vision and drive execution - Experience with pre-training and post-training techniques (self-supervised learning along with SFT/DPO/GRPO/PPO) - Self-starter with exceptional ownership mentality and ability to work in ambiguous environments - Passion for solving challenging problems and pushing the boundaries of AI research NICE-TO-HAVE - PhD in Machine Learning, Computer Science, or related areas - Experience with agent-based AI systems and multi-modal model development - Background in mid-training or pre-training of large language models - Publications in top-tier AI/ML conferences - Experience in fast-paced startup environments - Track record of translating research into production systems