Senior Backend Engineer - CRM

Job description

About this role

The CRM team owns everything that Close’s customers touch all day, every day: leads, contacts, opportunities, tasks, activities, custom objects, and search. Each time a call is logged, an email is threaded, a meeting is summarized, it lands on the record this team owns, and every agent we ship starts there. Today, this team is making big moves towards the next iteration of Close's core data model, our largest engineering effort currently underway. We're reopening assumptions made a decade ago and asking what a modern CRM's data model should look like — one that makes Close flexible for our small, scaling business customers. This is a long series of migrations and rearchitecture projects.

Responsibilities

• Rearchitect the core CRM data model. Rework how leads, contacts, opportunities, and custom objects relate to each other so each of our customers can represent their business the way it operates.
• Make Close usable by agents. The Agents team builds the MCP foundation; we build the tools and endpoints on top of it that expose the CRM's data and actions. The hard part is search: any MCP-compatible harness should be able to traverse, filter, and reason over a customer's CRM without learning our private query language first.
• Ship AI-native CRM features. This team built agentic enrichment: instead of paying a third-party data provider, an agent goes and researches whatever field you need, in bulk or automatically inside a workflow. There's a lot more of this ahead, including AI-powered data cleaning and richer AI search.
• Keep the hot paths fast. Finding a lead, updating a deal, scanning a pipeline, running a bulk action. Salespeople do these hundreds of times a week and no amount of clever architecture matters if they're slow.
• Tech you'll touch: Python, Flask, FastAPI, MongoDB, Elasticsearch, PostgreSQL, Redis, TaskTiger, MCP — plus whichever LLM ships next.

Qualifications and requirements

• A seasoned engineer. Python is our backbone, but perhaps you've worked across Go, Rust, or TypeScript. You pick the right tool for the workload rather than retreating to what you know.
• Architecturally minded. This is the thing we care about most. You've designed systems others built on, planned a migration that couldn't take the product down, and know how to sequence a rearchitecture into pieces that ship independently.
• Drawn to data model and retrieval problems. Schemas, relationships, normalization, query design, type systems. The plumbing decides whether a CRM feels fast and we pride ourselves on being fast.
• Experienced in search — this is a serious bonus. Elasticsearch especially, and anything adjacent: Lucene internals, query planning, relevance, indexing at scale, building or operating a search engine.
• AI-native in production. You've shipped meaningful LLM-backed or agentic features to real users, and you have a POV on where AI earns its place in a CRM workflow and where a structured system wins.