Job Specifications
What You'll Do
Build polished, fast product experiences in Next.js + TypeScript (design-quality UI, accessible components, responsive layouts).
Design clean backend contracts and APIs (FastAPI/Python) for workflows, jobs, billing, and auth.
Evolve our workflow engine (DAG/state machine): scheduling, retries/idempotency, rate limits, progress tracking, logs, and user-visible run history.
Ship data pipelines and background workers (Celery/queues), with strong observability and alerting.
Scale and harden the stack on AWS + Docker (ECS/EKS or similar), with caching (Redis), pagination, backpressure, and cost controls.
Implement product features end-to-end: DB schema - API - UI, including tests and telemetry.
Contribute to billing (Stripe), permissions/RBAC, and realtime UX (SSE/WebSockets) for run status.
Maintain a high bar for reliability, performance, and UX; instrument everything.
Minimum Requirements
Strong backend skills in Python (FastAPI preferred): API design, testing, error handling, schema/versioning.
Experience deploying/operating with Docker and a major cloud (AWS preferred: ECS/EKS/Lambda, SQS, CloudWatch).
Proven work on workflow/job systems: queues, retries, idempotency keys, scheduling, and run histories.
Comfortable with Postgres (or MySQL) schema design and Redis (caching, rate limiting, locks).
Performance mindset: RSC/Server Actions, data-fetching patterns, bundle/TTFB tuning, DB query perf.
Production observability: logs, traces, metrics (e.g., Langfuse, OpenTelemetry, Sentry), and on-call pragmatism.
Solid security hygiene (PII handling, secrets, JWT/session auth, CSRF/CORS, least-privilege access).
Nice to have
Next.js App Router, RSC, Server Actions, edge/runtime primitives (Vercel/AWS CloudFront).
Prisma/Drizzle; tRPC or GraphQL; Zod/JSON schema; feature flags/experimentation.
Realtime UX (WebSockets/SSE), optimistic updates, background revalidation.
Celery/RQ, event sourcing, or state-machine libs (e.g., Temporal, Dagster, Prefect, XState).
Stripe (subscriptions, metered usage, webhooks) and customer-visible billing pages.
CI/CD (GitHub Actions), IaC (Terraform), and basic cost/perf dashboards.
Search/retrieval familiarity (OpenSearch/Elasticsearch) and simple relevance tuning.
Testing: Playwright (e2e), Vitest/Jest, pytest; accessibility checks.