Backend Engineer Talent Network

$70–150/hrrate as listed

Mercor maintains a standing pool of backend engineers who write, review, and grade tasks used to train and evaluate frontier AI models on API design, databases, and distributed systems. You apply once, complete an AI screening interview, and get matched to contract projects as labs request backend expertise — typically 15–30 hours per week at observed rates of $70–150/hr.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work actually involves

This is not a software job in the usual sense — you are producing the material that teaches and tests models. Day-to-day work tends to fall into a few patterns depending on the project: authoring realistic backend tasks (a failing service, an N+1 query, a race condition in a job queue) with reference solutions and test cases; reviewing model-generated code and writing structured critiques of why an approach is wrong or merely mediocre; and ranking multiple model outputs against a rubric where the differences are subtle. Some projects lean toward multi-turn agentic evaluation, where you judge whether a model correctly navigated a real repository, read logs, and made a defensible change.

The hard part is rarely knowing the right answer. It is articulating why — in writing, precisely enough that another reviewer or a training pipeline can use it. Engineers who are fluent in tradeoff language (consistency vs. availability, normalization vs. read performance, retries vs. idempotency) do well here. Engineers who can only say "this looks fine" do not.

What the platform screens for

Mercor's process is application, credential verification, then an AI-led interview. The interview probes depth: expect to be asked about systems you personally built, then pushed on specifics — what the actual throughput was, why you chose that datastore, what broke in production and how you found it. Vague or resume-recited answers get filtered. The screen also assesses written communication and whether your stated availability is realistic, since project matching depends on it.

Verifiable professional backend experience — production systems, not coursework
Depth in at least one of: API/microservice design, database architecture and query optimization, distributed systems
Clear technical writing under time pressure

Logistics

Fully remote and largely asynchronous. Applying places you in a network, not a role — matching happens on a rolling basis when a lab requests backend expertise, and there can be a gap between approval and first invitation. Once matched, commitments typically run 15–30 hours weekly with flexible scheduling, though some projects impose review turnaround windows. The $70–150/hr band reflects observed rates across projects and varies with seniority, specialization, and the contracting lab; it is not a guarantee.

What they look for

Professional backend engineering experience with production systems
API design and microservice architecture
Database architecture, indexing, and query optimization
Distributed systems concepts (consistency, fault tolerance, queuing)
Clear written technical explanation and critique
Self-directed remote work with async communication

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you have at least 3 years of professional software engineering experience with primary responsibility for backend systems?
Professional backend experience · required
Have you designed, implemented, or significantly contributed to REST APIs, gRPC services, or microservice architectures in a production system?
API and microservices knowledge · required
Do you have reliable internet, a quiet workspace, and the ability to work independently on a flexible schedule from your home or office?
Remote work setup · preferred
Are you located in or able to work during US business hours (or state your primary working hours)?
Time zone · preferred

0 of 4 answered

Included with this listing

Interview prep — 10 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Whether you've genuinely architected systems (not just implemented features), and whether you think in tradeoffs between consistency, scalability, and maintainability.
A strong answer: Name a concrete system you owned, describe the API contract and data model you chose, explain one non-trivial choice (e.g., why you picked a particular database or communication pattern), and show you understood the constraints that drove it.
Where people lose it: Describing a system someone else built, or listing technologies without explaining why they were the right fit for that problem—this signals you follow tutorials rather than reason about architecture.