Member of Technical Staff (Software Engineer, Agent Capabilities)

Description

Core Tech Stack
Python, Go, PostgreSQL, DynamoDB, AWS, TypeScript, React

Experience Level
Senior (8+ years typical, exceptional candidates with less considered)

Job Type
Full-Time

Perplexity launched Computer in 2026, one of the defining products of the agentic AI era, and every action an agent takes inside it gets metered, budgeted, and settled through a billing platform this role helps own. You’ll join the Capabilities team, which sits right where frontier AI research meets product — turning breakthroughs in model behavior into reliable, production-grade features for both human users and AI agents.

About the Team

Millions of people already use Perplexity to turn knowledge into action. As new model capabilities emerge — better reasoning, planning, memory, and agent-to-agent collaboration — the Capabilities team is usually first to evaluate what’s possible, decide how it should show up in the product, and ship it as something real people and agents can rely on. This is a high-ownership role sitting across frontier AI research, agent systems, platform engineering, and product.

Why Perplexity

• Craftsmanship: building high-quality, tasteful products for both AI-native and AI-curious users.
• Ownership: you find the problem, design the solution, and ship it yourself.
• Entrepreneurship: thinking like founders, moving with urgency, hustling for each other and for users.
• Scholarship: working alongside strong peers in pursuit of real knowledge and truth.
• Partnership: amplifying each other’s strengths and breaking down silos.

What You’ll Do

• Dig into emerging frontier-model behavior and figure out how advances in reasoning, planning, memory, and agent collaboration should show up in the product.
• Build and evolve Skills, Workflows, and Artifacts into one coherent system, shaping the architecture and product experience that lets both users and agents tackle increasingly complex tasks.
• Own capabilities end-to-end — from the user-facing interface down through backend services, evaluation, data models, and how it runs in production.
• Build scalable platform infrastructure that powers how capabilities get executed, shared, and managed across huge volumes of user and agent activity.
• Work closely with PM, Design, Data Science, Sales, and Research to spot high-impact opportunities in both consumer and enterprise experiences, validate new capabilities, and turn complex agent behavior into something simple and reliable.
• Set technical direction on genuinely ambiguous problems, and raise the bar for the team through design reviews and mentorship.

What You’ll Need

• Typically 8+ years of professional software engineering experience, with a real track record of owning complex products or systems from idea to production (strong candidates with less experience but a great track record are still encouraged to apply).
• Solid full-stack fundamentals — building user-facing products, backend services, and distributed systems that handle real scale.
• Strong product instincts: you can turn a vague need into something simple and reliable, and you ship without waiting for a perfect spec.
• Comfort making data-informed calls — you define the metrics that prove something works, then keep iterating on them.
• Experience designing abstractions or reusable platforms that let other engineers, users, or automated systems get complex things done.
• Real excitement about frontier AI, agent systems, and rapidly evaluating and shipping new model behaviors.

Nice to Have

• Experience building agentic systems — tool calling, subagents, long-running or autonomous task execution.
• Experience building developer platforms or reusable primitives like SDKs, plugin systems, or workflow engines.
• Background in evaluation, benchmarking, or quality systems for ML/LLM-powered products.
• Time spent at a fast-growing startup or on a high-ownership engineering team.