Search ML Engineer Intern: Build Retrieval Ranking

Perplexity AI; Greater London; On-site; GBP 17,000 - 22,000; Full time

Job summary

Perplexity AI in London is seeking a Search ML Engineer Intern to help build the next generation of search technologies, focusing on retrieval and ranking. You will work closely with experienced engineers to improve search quality and ship features that directly impact how users search and discover information.

The internship program runs 12–24 weeks full-time, in-person at the London office. You will design and implement components of the search platform, train evaluation models (LLM-based

Qualifications

• Strong foundation in ML and statistics with information retrieval or ranking coursework.
• Experience evaluating models using offline metrics.
• Familiarity with A/B testing is a plus.
• Self-driven, ownership and ability to work in a fast-paced environment.
• Internship duration: 12–24 weeks, full-time, in London.

Responsibilities

• Design and implement components of the search platform and model stack (retrieval, ranking, classification).
• Train and evaluate models for retrieval, ranking, and classification tasks (including LLM-based approaches).
• Support deployment and monitoring of search and ranking models in production.
• Help build and iterate on RAG pipelines for grounding and answer generation.
• Collaborate with Data, AI, Infrastructure and Product teams to deliver improvements quickly.

Skills

Python

ML frameworks

Information retrieval

Ranking

Recommender systems

Rust

Tools

PyTorch

TensorFlow

JAX

Job description

Perplexity AI in London is seeking a Search ML Engineer Intern to help build the next generation of search technologies, focusing on retrieval and ranking. You will work closely with experienced engineers to improve search quality and ship features that directly impact how users search and discover information.

The internship program runs 12–24 weeks full-time, in-person at the London office. You will design and implement components of the search platform, train evaluation models (LLM-based