AI Engineer - Personalization Infrastructure

Perplexity is seeking experienced AI Infrastructure Engineers to build the next generation of personalization infrastructure. In this role, you will develop cutting-edge systems that give our AI the ability to remember contexts, understand user intents, and deliver highly personalized experiences.

Responsibilities

• Architect and ship the next generation of personalization infrastructure: long‑lived LLM memory, long‑context understanding, conversation retrieval, async task execution, and related query retrieval
• Build event-to-embedding pipelines and feature stores; design vector and hybrid indices; implement retrieval/ranking services that power real‑time, personalized experiences.
• Partner cross‑functionally with Product and Research to translate signals into durable memory and measurable relevance lift; instrument metrics, run A/B experiments, and iterate quickly.
• Drive end‑to‑end ownership: data modeling, schema and contracts, batch/stream processing, model training, deployment, observability, and on‑call for owned services.

Qualifications

• Strong programming skills in production systems with hands‑on experience across data, serving, and infra layers.
• Proven experience building large‑scale retrieval and indexing infrastructure (vector/hybrid search, inverted indexes).
• Deep familiarity with LLM‑centric IR: RAG architectures, semantic retrieval, prompt/context optimization, and long‑context strategies.
• Experience building and deploying ML for personalization: retrieval/ranking models, feature engineering, offline evaluation, and online experimentation (A/B, interleaving).
• Ownership mindset, bias to action, and ability to deliver in fast‑moving, ambiguous environments.
• 4+ years of industry experience in distributed systems, search/recommendation, or ML systems. Experience with LLM memory or conversation retrieval is a strong plus.

The cash compensation range for this role is $200,000 - $280,000.