Software Engineer, Enterprise Knowledge Platform
PerplexitySan Francisco, CAposted 4h ago
- PayUSD / yr
- $220k – 405k
- Where
- San Francisco, CA
- Posted
- Oct 21
- 4 hours ago
- Type
- Full-time
- Setup
- On-site
At a glance
Build the backend and product platform that lets enterprise AI agents discover tools, understand organizational context, and execute permission-aware actions reliably across customer systems.
Summarized by AI from the original posting
What you'll do
- Own the connector runtime for native connectors, hosted MCP servers, and CLI-backed tools
- Build semantic and knowledge layers for tools, entities, terminology, and permissions
- Design discovery, search, and recommendation systems
- Build marketplace and integration activation experiences
- Improve connector reliability, observability, retries, idempotency, pagination, and recovery
- Implement authentication, credential isolation, and permission enforcement
- Build onboarding and publishing workflows for internal teams and external partners
- Own reliability targets, production monitoring, and incident response
Requirements
- 01Track record owning complex backend systems from design through production operation
- 02Strong system design skills
- 03Experience integrating heterogeneous APIs and services
- 04Hands-on cloud infrastructure experience with AWS and Kubernetes
- 05Deep expertise in connector runtimes, API platforms, MCP servers, semantic modeling, knowledge systems, search, ranking, recommendation, marketplaces, or developer platforms
- 06Ability to work across backend infrastructure and customer-facing product experiences
- 07Technical judgment in security-sensitive distributed systems
- 08Track record leading ambiguous cross-team initiatives and mentoring engineers
Skills
Full description
About the Role
The Enterprise Knowledge Platform team builds the data and action layer that connects Perplexity's agents to the systems where enterprise work happens. Useful agents need more than access to APIs: they need to discover the right tools, understand an organization's data and terminology, and act reliably within its permissions. This team brings native connectors, MCP servers, and CLI-backed tools into a consistent platform, and builds the marketplace through which customers adopt integrations and partners extend the ecosystem.
This is a product and platform engineering role with end-to-end ownership, from designing connector infrastructure and semantic models to shipping the experiences customers use to discover, connect, and activate integrations. You'll build the runtime that executes tool calls, the knowledge layer that makes enterprise context reusable across connectors, and the developer and marketplace foundations that help the ecosystem grow. Working with AI Research and Enterprise teams, you'll turn fragmented systems into reliable, permission-aware capabilities that agents can use to complete real work. The impact is direct: every improvement helps customers put their existing software and knowledge to work through Perplexity.
Key Responsibilities
Own the connector runtime that registers, hosts, and executes native connectors, hosted MCP servers, and CLI-backed tools through a consistent agent-facing interface.
Build the semantic and knowledge layer, including tool and entity schemas, capability metadata, relationships, organization-specific terminology, and reusable corrections, while preserving provenance and source-system permissions.
Design discovery, search, and recommendation systems that help agents find relevant tools and context and help users and organizations choose integrations that fit their work.
Build the marketplace and activation experience end to end, from browsing and recommendations to installation, configuration, admin enablement, and ongoing adoption.
Make connector execution reliable and observable, with structured results, retries, idempotency, pagination, rate-limit handling, and recovery from partial failures. Own reliability targets, production monitoring, and incident response.
Implement connector authentication, credential isolation, and permission enforcement, partnering with Trust & Governance, Security, and Backend Platform on shared policies and tenant boundaries.
Build the onboarding and publishing path for internal teams and external partners, including schemas, test fixtures, validation, review, and versioning.
Qualifications
4+ years of software engineering experience, with a track record of owning complex backend systems from design through production operation.
Strong system design skills, including building reliable, scalable platforms that integrate heterogeneous APIs and services with clear security and isolation boundaries.
Strong proficiency in a backend language such as Python, Go, or Rust, and hands-on experience with cloud infrastructure such as AWS and Kubernetes.
Deep expertise in at least one of the following:
Connector runtimes, API platforms, or MCP servers, including authentication and authorization.
Schema design, semantic modeling, or knowledge systems that capture relationships and organizational context.
Search, ranking, recommendation, or tool discovery systems.
Marketplaces, developer platforms, or enterprise onboarding and activation experiences.
Ability to work across backend infrastructure and customer-facing product experiences, translating complex integration requirements into clear APIs, developer tools, and intuitive workflows.
Sound technical judgment in security-sensitive, distributed systems, balancing credential isolation, permission correctness, and failure behavior with performance, usability, and delivery speed.
A track record of leading ambiguous, cross-team initiatives from architecture through adoption, setting technical direction for shared platforms, and raising engineering quality through mentorship.
If you're excited about this role, we encourage you to apply even if you don't meet every qualification. We value demonstrated impact, deep ownership, and the initiative to turn ambiguity into results.
Compensation: $220K - $405K