Software Engineer, Infrastructure (Early Career)

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 If you want to be part of the AI wave and are drawn to the systems challenges of a product used by both people and agents, this role is for you. Lightfield's infrastructure supports human-interactive requests, agent-driven queries and actions, scheduled agents, and continuous ingestion of email, calendar, and meeting data. Based in our Cambridge office, you'll work on the databases, scheduling, permissions, and distributed systems that keep these workloads fast, isolated, and reliable as they share the same underlying customer data. The work is not research, but many of the same habits matter: forming hypotheses, reading unfamiliar systems deeply, designing careful experiments, and explaining what the evidence shows. You'll use those habits on problems like these: - Querying a schema-flexible data model. Customers define their own object types, fields, and relationships at runtime, while the underlying data is stored in Postgres. We work on query planning, indexing, and data-access patterns that keep those records fast to filter and aggregate as datasets grow. - Record-level permissions across relationships. Access can come from ownership, direct grants, group membership, or inheritance from another record. We need a permissions model that evaluates those paths efficiently and produces the same answer across the application, API, search, dashboards, and agents. - Balancing real-time and batch workloads. Search indexing competes with historical imports and backfills for shared database and queue capacity. We use concurrency controls, prioritization, and backpressure to keep real-time updates moving while still processing large batches. You'll design and review systems, debug production behavior, and gradually take full ownership of one or more subsystems. You'll ship production work early with close design and code review, then take on increasingly open-ended projects as you demonstrate readiness. You'd join a small, systems-focused engineering group in our new Kendall Square office early enough to help establish how it works, alongside a site lead who is one of our most senior infrastructure engineers. WHAT YOU'LL DO - Ship production work early with close design and code review, then grow into end-to-end ownership of a system. - Investigate systems methodically: reproduce behavior, add instrumentation, form a hypothesis, and test it against production evidence. - Work on query performance, data modeling, permissions, scheduling, and keeping customer-facing data fresh. - Debug real production issues across application code, databases, queues, and infrastructure. - Learn how production systems are designed and operated through code review, incident response, and owning what you ship. WHAT WE'RE LOOKING FOR - Strong computer science fundamentals and a degree in CS or a related field. - Depth somewhere systems-shaped: research experience, coursework you went deep on (operating systems, databases, distributed systems, security), a systems-flav