Sofware Engineer, Backend (Staff)

ABOUT LIGHTFIELD Lightfield is an AI-native CRM that assembles itself from your email, calendar, and meetings. It captures every interaction and turns it into organized context: accounts, tasks, follow-ups, and insights, so nothing slips through the cracks. We’re rethinking CRM from first principles. Instead of forcing teams to maintain rigid systems, Lightfield learns from how companies actually work, adapting, automating, and surfacing the insight that drives growth. We’re building the CRM platform we always wished existed: fast, intelligent, and genuinely helpful. The market agrees. Over 5,000 companies have used Lightfield since we launched in November, making us the fastest-adopted product in our category. We are backed by Greylock, Lightspeed, and Coatue, and our founders previously built Tome, a generative AI presentation product used by over 25 million people. Before Lightfield, our team worked on Llama, Instagram, Facebook Messenger, Pinterest, Google, and Salesforce. ABOUT THE ROLE At Lightfield, the backend is not a layer beneath the product. It is the product's foundation. Lightfield builds a living model of a company's customers from emails, calls, meetings, and other signals. It structures that raw history into customer-defined objects, attributes, relationships, and versioned values, then exposes the same underlying capabilities to the product UI, external APIs, automations, and agents. Decisions about the data model, authorization, event propagation, and execution semantics directly determine what customers can do. As a Software Engineer, Backend, you'll own those decisions for important product capabilities. You'll start with an ambiguous customer or product problem, define the behavior and system boundaries, design the data model and APIs, sequence migrations, and carry the capability through rollout and production. Recent examples include record-level permissions, field history and audit, durable workflows, and consumption metering. These foundations already operate at meaningful scale: - Our Postgres fleet serves 5B+ queries a month — roughly 2,500 QPS steady state, with sustained spikes past 25,000 QPS — and database workload more than doubled last month. - Redis sustains ~50,000 commands per second behind a job platform that executes 10M+ background job runs a day across ~170 queues. - We ingest tens of millions of emails and calendar events a month. Scale makes the product work technically serious: new abstractions must be fast, reliable, observable, and safe under real load. But this is not primarily a role for keeping the lights on. It's a chance to define foundational pieces of an AI-native product while the architecture is still taking shape—and to see those decisions change what customers and agents can accomplish. This role can be based in San Francisco or Cambridge. In San Francisco, you'd work from our HQ alongside the founders and most of the engineering team. In Cambridge, you'd join an initial group of staff-level engineers at our new Kendall Square site, working alongside one of our most senior infrastructure engineers. We aim to build the site and organization around this group as the company scales. WHAT YOU'LL DO - Own core product capabilities end to end—from customer requirement and behavioral semantics through data model, API, migration, rollout, and operation. - Design the shared abstractions used by Lightfield's UI, public API, agents, automations, and internal engineering teams. - Evolve a schema-flexible, graph-shaped data model on Postgres, where customers define objects, attributes, and relationships at runtime and values change over time. - Build systems for authorization, history, event propagation, durable execution, and metering that become visible parts of the customer experience. - Partner directly with product engineers, product leadership, and customers to turn ambiguous needs into practical backend architecture. - Make high-throughput Node.js services, Postgres query path