SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Lightfield is an AI-native CRM that automatically assembles customer relationship data from email, calendar, and meetings. The company is backed by Greylock, Lightspeed, and Coatue, with founders who previously built Tome (25M+ users) and team members from Llama, Instagram, Facebook, Pinterest, Google, and Salesforce.
You'll build the next generation of data infrastructure for a fast-growing AI product serving thousands of customers. The current system handles 5B+ Postgres queries monthly (~2,500 QPS steady state, spikes to 25,000 QPS), 50,000 Redis commands per second, and ingests tens of millions of emails and calendar events monthly. Database workload more than doubled last month, creating scaling pressure across backend, infrastructure, and data systems.
The role spans data infrastructure but welcomes work across backend systems, infrastructure, and product-facing data problems. You'll evolve a deliberately simple, pragmatic stack—Postgres as system of record, sharded transactional outbox for change events, Redis-buffered sync into Typesense for search, BullMQ for processing, and Postgres-backed customer-facing analytics with row-level security—into best-practice data architecture.
Key responsibilities include: scaling the analytics engine behind customer-facing dashboards (query performance, row-level security, workload isolation, observability); designing ingestion paths, event models, schemas, and query patterns with freshness and correctness guarantees; evolving the schema-flexible, graph-shaped data model (entity-attribute-value with typed edges, versioned attributes, relationship history) to remain fast as customer data grows; building historical reporting and auditability foundations; creating reliable data systems for usage metering, pipeline generation, and AI evaluation; and setting technical direction for data systems as the company scales.
First-year projects include scaling analytics serving, zero-downtime schema migrations for an 18-collection Typesense deployment, a usage-metering pipeline for consumption billing, historical and audit data modeling, and evaluation data infrastructure for AI agents. You'll own technical direction for how data is modeled, moved, and served across Lightfield, and shape the architecture, abstractions, and team as the company scales.
The role can be based in San Francisco (HQ, alongside founders and most engineering) or Cambridge (new Kendall Square site, infrastructure-focused, alongside senior infrastructure engineers). This is data infrastructure work, not BI or dashboarding—a fit for someone who likes high-volume data systems, pragmatic architecture decisions, and building foundations that product and engineering teams depend on.
