Founding Software Engineer

Senior-level engineers with 4+ years of experience building production-grade backend systems, particularly in JavaScript/TypeScript, Node.js, and NestJS, are best suited for this founding role.

They should be skilled in designing REST and GraphQL APIs, working with PostgreSQL and Redis, implementing authentication and authorization, and using Docker, CI/CD, and Git.

The ideal candidate is familiar with owning architecture and backend performance, reliability, observability, and integrations involving AI pipelines, third-party services, async workers, or event-driven systems in a fast-paced startup.

How would you structure a scalable NestJS backend for AI-driven recruiting workflows?

I would organize the application into modular domains such as candidates, jobs, outreach, integrations, and authentication, with clear boundaries between controllers, services, repositories, and workers. I would use dependency injection, shared infrastructure modules, and asynchronous processing to keep the system maintainable as features and teams grow.

How would you design REST and GraphQL APIs that serve both frontend applications and AI services?

I would use REST for resource-oriented operations and external integrations, while using GraphQL where clients need flexible, aggregated views. Both interfaces should share domain services, enforce consistent authorization and validation, and include versioning, pagination, rate limiting, and clear API documentation.

How would you model recruiting data in PostgreSQL?

I would normalize core entities such as organizations, users, candidates, jobs, applications, and interactions, while adding carefully chosen denormalized fields for high-volume reads. I would define foreign keys, indexes, constraints, and migration practices explicitly, and use transactions for workflows that update multiple related records.

How would you use Redis in this platform?

Redis could support short-lived caching, distributed locks, rate limiting, session or token metadata, and queues for asynchronous work. Cached data should have explicit TTLs and invalidation rules, and the system should remain correct if Redis is unavailable or contains stale values.

How would you implement reliable background processing for AI pipelines and third-party integrations?

I would use a queue system such as BullMQ with durable jobs, retries using exponential backoff, idempotency keys, timeouts, and dead-letter handling. Workers should expose job metrics and preserve enough state to resume or safely retry operations without duplicating outreach or external side effects.

How would you design authentication and authorization for a multi-tenant B2B SaaS platform?

I would separate authentication from authorization, supporting OAuth2 or SSO where required and issuing securely managed sessions or tokens. Authorization should enforce tenant isolation and role- or permission-based access at the service layer, with audit logging for sensitive actions and defense against insecure direct object references.

What strategies would you use to improve backend performance and reliability?

I would measure first using profiling, database query analysis, latency percentiles, and tracing. Common improvements include proper indexes, bounded queries, caching, connection pooling, asynchronous work, horizontal scaling, timeouts, circuit breakers, and graceful degradation for dependent AI or third-party services.

How would you make the system observable in production?

I would implement structured logs with correlation and tenant-safe context, metrics for request latency, errors, queue depth, database health, and external dependency performance, plus distributed tracing across API and worker flows. Alerts should be tied to user-impacting symptoms and supported by actionable dashboards and runbooks.

How would you safely integrate AI inference services into backend workflows?

I would isolate inference behind a well-defined service interface with request validation, timeouts, retries, versioned prompts or models, and usage limits. Results should be persisted with provenance and status, while sensitive candidate data must be protected through access controls, encryption, redaction, and provider-specific privacy policies.

What would a robust Docker and CI/CD workflow look like for a TypeScript backend?

I would use multi-stage Docker builds, pinned dependencies, non-root containers, health checks, and environment-specific configuration managed outside the image. CI should run formatting, linting, type checks, unit and integration tests, security checks, database migration validation, and deployment gates before promoting artifacts through environments.

Neuroscale is building Arbi, an AI recruiting operating system that unifies talent workflows and orchestrates intelligent agents from sourcing through hiring. As a remote Founding Software Engineer, you’ll design scalable backend services, REST and GraphQL APIs, authentication systems, and infrastructure while collaborating with AI, frontend, and product teams. The stack includes TypeScript, Node.js, NestJS, PostgreSQL, Redis, Docker, CI/CD, and Git. With a $100K–$200K salary, meaningful early-stage equity, flexible work-from-anywhere arrangements, learning support, and a path to Engineering Lead, this is a high-ownership opportunity to shape a fast-growing company from the ground up.