Senior Software Engineer - Onboarding Automation
at Loancrate
- Seniority
- Lead Manager
- Employment
- Full Time
- Location
- United States
- Posted
- 5d ago
typescriptreactexpressgraphqlpostgreselasticsearchkafkaawsdockerterraformpulumici/cd
at Loancrate
WHAT IS LOANCRATE? We started Loancrate to make home-buying simpler and less expensive for lenders and borrowers (us!). Today, mortgage lenders are stuck running their companies on software products built 20 years ago. These products are slow, unstable, and don't lead to material improvements in efficiency. When using these systems, the average human cost to originate a loan is still over $11,000. Loancrate builds AI-native tooling to automate mortgage workflows. Our ultimate goal is fully automated origination, which has the potential to save lenders over $16B in operating expense per year. Since starting in 2020, our remote team has enabled our customers to power >$85 billion in new home loans. We are a group of people excited to tackle the complexity of the home-lending industry. We care about collaboration, very open communication covering the good & the bad so that we learn from our decisions quickly, and ultimately having fun while we're building. You'll fit in well if you like diving deep quickly! THE OPPORTUNITY Our enterprise pipeline is outrunning our implementation capacity. The number that caps our growth isn't sales — it's how long it takes to get a new lender live. Onboarding a large mortgage lender means standing up a tenant, modeling their org and branch hierarchy, encoding their workflows and approval rules, wiring dozens of vendor integrations under their credentials, migrating an in-flight pipeline off a legacy system, and proving all of it in UAT before a single real borrower touches it. In this industry that work is measured in quarters, and most of it is done by hand. We're looking for an engineer to turn that from a services project into a product. Every enterprise onboarding should look less like a bespoke implementation and more like running a pipeline: declarative config, automated validation, reproducible environments, and a clear diff between what a customer asked for and what's actually running. Get this right and our sales ceiling stops being an implementation-capacity problem. WHAT TO EXPECT You'll continue our team tradition by shipping code to production on your first day! Following that, within your first month, you'll dive into your first mission-critical project, such as... - Building tenant provisioning that stands up a fully configured enterprise environment — org hierarchy, roles and permissions, workflows, document types, integration credentials — from a versioned config file, in hours instead of weeks. - Designing configuration-as-code for customer setup: diff it, review it in a PR, promote it from sandbox to UAT to production, and roll it back when a change goes sideways at 6pm on a funding day. - Building the migration toolchain that pulls an in-flight loan pipeline and historical data off Encompass and other legacy systems, then produces a reconciliation report the customer's ops leadership will actually sign off on. - Extending our release machinery for a multi-tenant world: per-tenant feature flags, staged rollouts by lender, and deploys that don't require a human to reason about which customer is on which configuration. - Building the smoke tests and conformance suites that let an implementation lead prove a customer's integrations work end to end without pulling an engineer into the room. TECH STACK Nearly all of our code is written in TypeScript. We use React on the frontend, Node/Express on the backend (running inside Docker), GraphQL/Apollo on both sides, and Postgres as our database. We use AWS (ECS, RDS, ElastiCache, Kafka, ElasticSearch, and more) and manage all of our infrastructure with Terraform. We believe that high engineering velocity drives high product velocity (and, in turn, our success as a company). Keeping this in mind, we've invested heavily in creating a great developer experience that enables engineers to ship fast. Some highlights of our stack: - End-to-End Type Safety - We generate TypeScript types from our Postgres and GraphQL schemas so that when we change our database schema we see type errors in any incompatible frontend React components. - Full-Stack Code Sharing - Our build tooling supports seamless sharing of TypeScript files between the client, server, and other packages in our mono-repo so that we never have to define functions or configuration values twice. - Reproducible Infrastructure - We've used Terraform to manage 100% of our infrastructure since day 1 and have configured our CI pipeline to automatically deploy temporary staging environments for every pull request. - First Class Testing Utilities - We have a comprehensive library of utilities that mock production-like data and feed automated end-to-end, integration, and performance tests. Our wide-ranging test suite allows us to move fast while maintaining nearly perfect uptime. The tooling you build will sit on top of all of this — and where it doesn't fit, you'll change the foundation rather than work around it. WHAT WE REQUIRE We'll be direct about the bar. We're looking for evidence of all three: 🏭 You've productized an implementation process before. You've watched a team do the same painful setup for the fifth customer and replaced it with tooling — and you know the difference between a config system that scales and one that becomes a second product nobody maintains. 🧾 Configuration-as-code instincts. Versioning, diffing, validating, and promoting customer configuration through environments. Bonus if you've dealt with the schema-migration problem that shows up when config outlives the code that reads it. 🚚 Data migration under pressure. Moving real production records off a legacy system with validation, reconciliation, and a rollback plan. Mortgage pipelines are live money; a bad migration is not a Jira ticket. NICE TO HAVE (It's okay not to have all of these things - these are just some skills we are excited about!) 🏢 Multi-tenant experience. Per-tenant isolation, per-tenant rollout, and the operati