ZoomInfo's Core Data team builds the company and contact data behind the Go-To-Market Intelligence Platform: hundreds of millions of records, resolved to the right entity, kept accurate, and delivered to more than 35,000 customers.
Core Data is moving to AI-native data management, where an agent pursues the evidence for an entity, decides every attribute, and records why.
Responsibilities
Run the benchmark and golden-gate system: move judge runs, golden-gate results, arms scorecards, and cost per record to a scheduled system against staging
Complete the entity-agent pipeline: enricher fan-out, location sets, hierarchy batch files; migrate classifiers to the shared runner
Build the evidence adapters: registries, live web, mail-tenant signals, external research vendors behind one envelope with a replayable vendor store and per-vendor kill switch
Deploy onto the ZoomInfo Agentic Platform: carry company and contact agents through the production plan onto ZAP, build the ingest adapter side, keep a rollback path per version
Make the loop self-service: build review queue, tagging tools, blast-radius viewer, and batch runners
Keep the agent surfaces alive: own health and deployment of company classifier, contact classifier, agents hub, and judge platform
Instrument everything: cost per call, per row, per stage; quality per version; alerts before a human notices
Requirements
5+ years of software engineering experience, including at least 2 years shipping LLM-based systems to production: agent loops, tool use, retrieval, structured output, and evaluation harness
Strong Python and TypeScript, comfort across an API layer, a batch runner, and a small web UI
Experience building platform others operate: adapters, runners, queues, or internal tools designed for replay, idempotency, and observability
Working knowledge of data pipelines and entity resolution
A track record of taking a prototype to production, treating test sets and benchmarks as the contract
Instrumentation by default: know the cost and latency budget of every system shipped
Strong bias for action in a fast-moving, distributed organization
Preferred
Experience with B2B company or contact data, identity or registry data, or entity resolution at scale