Role at a Glance
Perplexity is hiring a Member of Technical Staff (AI Software Engineer, Agents) in San Francisco to build the scaffolding that gives frontier AI models real-world reach. Core language is Python; TypeScript, Go, and Rust are bonuses. The role requires demonstrable depth in at least one of three tracks: context engineering, post-training/RL, or browser automation. Compensation: $220,000–$405,000/yr. No remote policy is stated in the listing.
About the Role
This is a high-ownership engineering position on Perplexity's Agents team — a cross-functional group of backend, full-stack, and AI/ML engineers who build the infrastructure powering the company's agentic products. The work sits squarely at the intersection of applied AI research and product engineering: you will architect the connective tissue between frontier language models and the tools, browsers, and environments those models need to take meaningful action in the world. Scope spans model integration, browser automation, post-training pipelines, and user experience quality — across desktop, mobile, and headless cloud contexts — and the team's systems serve millions of active users.
About Perplexity
Perplexity is an AI-powered search and answer platform that has expanded well beyond its original Q&A format into full agentic task execution. The Agents team is specifically responsible for building general-purpose AI systems capable of navigating digital environments autonomously on a user's behalf — a meaningfully harder problem than retrieval or summarization. Their current product surface includes a generalized intelligence platform called Perplexity Computer, the Comet ecosystem, and a developer-facing Agent API. The team is deliberately small and execution-focused, with engineers expected to own large product areas rather than narrow feature slices.
What You'll Do
- Design and build agent harnesses that connect powerful AI models to the tools, APIs, and environments needed to complete economically meaningful tasks for users.
- Drive AI capability advances across multiple product layers in the agents space — from model integration to user-facing surfaces.
- Leverage and extend state-of-the-art models, cloud infrastructure, and browser technologies to push the frontier of what agents can accomplish, then scale those capabilities to a rapidly growing user base.
- Maintain a high quality bar in both AI agent performance and end-user experience — the role treats these as equally important, not a trade-off.
- Partner closely with engineers, designers, product managers, and data scientists to wire Perplexity's core platform capabilities into new agentic products, and vice versa.
- Own product reliability, code quality, AI evaluation frameworks, testing infrastructure, and ongoing maintenance work across the broader team.
Must-Have Qualifications
- Thorough, hands-on familiarity with the full AI product stack — from training and inference through to deployment and user-facing integration.
- Proficiency in Python as a primary working language.
- Significant, demonstrable experience in at least one of the following specialization tracks: (1) context engineering and tool-call interface design for large frontier models; (2) post-training and reinforcement learning, particularly for multimodal model architectures; or (3) browser technologies including Chrome DevTools Protocol (CDP), Playwright, or browser extension development.
- Strong product sense — genuine care for how users experience the product, not only technical correctness.
- Ability to thrive on a compact, fast-moving team; must be willing to take full ownership and drive work end-to-end without heavy scaffolding.
Bonus Skills
- Working proficiency in TypeScript, Go, and/or Rust in addition to Python.
Skills Breakdown: Required vs. Bonus
Python is the stated hard requirement — it gates the role. TypeScript, Go, and Rust are explicitly flagged as bonuses, meaning cross-language breadth is valued but not a filter. The specialization requirement is pick-one-of-three: context engineering, post-training/RL, or browser automation. This T-shaped profile — broad AI product literacy plus deep expertise in a single track — is characteristic of staff-level AI engineering roles. At MTS level, Perplexity expects engineers to scope and ship complex systems independently, so end-to-end ownership instincts and strong product intuition carry real weight alongside the technical depth requirement.
Compensation
The employer's posted salary range is $220,000–$405,000 per year. This is an exceptionally wide band — nearly a 2× spread — which is typical for MTS roles at growth-stage AI companies, where candidate seniority, prior ownership scope, and equity negotiation all significantly influence placement within the range. The job posting does not detail equity grants, annual bonus, or benefits components, so total compensation may differ substantially from base salary alone. The salary figure above comes directly from the employer's listing; JobForMe does not estimate or modify compensation data. Candidates should ask about the full package structure early in the recruiter conversation.
Location & Work Arrangement
This role is based in San Francisco, California. The posting does not specify a remote or hybrid work policy — candidates should confirm working arrangements directly with Perplexity's recruiting team before applying. No visa sponsorship details are included in the listing; candidates who require sponsorship or work authorization support should raise this question during initial outreach.