About
pdfFiller is an online platform for creating, editing, and signing PDF documents. It enables over 10 million users each month to create and edit PDFs, send them for signing, and securely store documents in one application. The team builds an AI assistant that understands and acts on users' documents, an MCP server that exposes pdfFiller's capabilities to external AI clients, and AI-powered improvements to existing product features.
Responsibilities
•Own features end to end, from requirements to production and post-release monitoring of edge cases
•Own the automated quality of delivered work, including unit and integration tests and evaluation cases for LLM behavior
•Treat prompt regressions, evaluations, and non-determinism as everyday engineering concerns
•Review code and documentation to maintain codebase consistency
•Monitor production and respond to incidents as part of the duty rotation
•Propose improvements to processes, tooling, and legacy code as part of the team's ongoing modernization
•Build tool-calling, streaming, and RAG features, owning them from problem definition to production metrics
•Work with the specified backend, frontend, and infrastructure stack
Requirements
•Experience with LangGraph or LangChain, agent orchestration, and state machines
•Experience with RAG and vector search using pgvector
•Familiarity with SSE or other streaming protocols
•Experience with observability tools such as OpenTelemetry and the Grafana stack
•Experience with LLM evaluation, including evaluation datasets and LLM-as-judge
•Experience writing technical documentation, including C4 and ADR
Nice to have
•Experience with LangGraph or LangChain, agent orchestration, and state machines
•Experience with RAG and vector search (pgvector)
•Familiarity with SSE or other streaming protocols
•Experience with OpenTelemetry and the Grafana stack
•Experience with LLM evaluation and evaluation datasets
•Experience writing technical documentation (C4 model, ADRs)
Conditions
•Educational budget
•Regular knowledge-sharing sessions
•Clear growth path
