Backend / AI Systems Engineer

Spellbook

9d•CA$140,000 - CA$210,000•Remote

About The Position

Spellbook is the most comprehensive AI copilot for transactional lawyers. It works directly inside Microsoft Word to help legal teams draft, review, and negotiate contracts up to 10x faster and with greater precision. Today, more than 4,000 law firms, in-house teams, and solo practitioners rely on Spellbook to simplify their workflows and eliminate the drudgery of everyday contract work. We are backed by leading investors including Khosla Ventures, Thomson Reuters Ventures, Inovia Capital, The LegalTech Fund, Bling Capital, and Moxxie Ventures. The company recently raised $50 million in Series B funding, led by Keith Rabois at Khosla Ventures, bringing its total funding to more than $80 million. This is an existing vacancy. You’ll build the backend systems that power Spellbook’s product experiences, including both traditional software services and AI-driven workflows. This is a hands-on, high-ownership role for an engineer who cares about product outcomes and understands what requires scale vs what does not, and can turn ambiguity into robust systems that lawyers trust.

Requirements

You have 5+ years of experience building backend systems in production.
You have strong fundamentals in backend engineering and distributed systems: APIs, data modeling, concurrency, queues, AI inference
You have strong NoSQL (ideally MongoDB) data modeling and performance tuning: schema design/denormalization tradeoffs, indexing strategy, query optimization, and profiling/diagnosing bottlenecks in production.
You can take an ambiguous, high-impact problem from “we should do something here” to a clear plan and shipped outcome.
You have strong communication skills. You can write and explain technical decisions clearly to engineers and non-engineers.
You are a self-starter and problem-solver motivated by curiosity and a desire to help others succeed, encouraged by continuous improvement.
You are pragmatic and understand that not everything we ship needs to scale, but know where to draw the line
You are a team player who is motivated to help Spellbook succeed. When things break you are eager and able to help fix them. You think of and implement ways to help and improve the work of the team as a whole.

Nice To Haves

Experience building LLM-powered systems in production, including prompt iteration, tool/function calling, retrieval patterns, and agent orchestration.
Experience operating high-scale retrieval systems (vector + lexical search) and measuring/tuning retrieval quality and latency in production.
Experience building evaluation frameworks (quality metrics, dataset curation, regression testing).
Experience with AWS CDK and operating production services in AWS.

Responsibilities

Build and scale backend systems that support core product functionality across Spellbook (Reviews, Chat, Library, and more).
Design low-latency, high-reliability search, inference, and orchestration layers for AI features
Build and operate RAG retrieval at scale with production-grade performance optimization and permissions-correct data isolation.
Own core platform concerns for AI: rate limiting, retries, consistency, fallbacks, safe degradation
Partner closely with Product and Design to make good tradeoffs between latency, accuracy, UX, and reliability.
Use modern development workflows, including agent-assisted coding, to accelerate delivery while maintaining rigorous review, testing, and security standards.
Participate in on-call and incident response as needed, and improve feature reliability in the process