About Andarz
Andarz is building an AI learning companion that delivers useful, standardized answers (3-2-1 + mini-quiz) fast. Backend is the engine that makes it real.
Why this role?
You’ll architect and ship the APIs and services that power chat, retrieval, evaluation, and payments—high impact, user-visible, and performance-critical.
What you’ll do
- Design, build, and operate scalable services (auth, chat/session, Q&A workflow, retrieval/RAG gateways, quiz, feedback/likes).
- Ship clean, well-tested APIs (REST/gRPC) with observability (logs/metrics/traces) and SLOs (e.g., p50/p95 latency, uptime).
- Integrate with MLOps for model routing, evaluation, and guardrails.
- Own data access patterns (Postgres/Redis), queues, and job workers.
- Implement secure payments/credits (“Endo” currency), rate limits, and abuse protection.
- Partner with Frontend/PM/Design to deliver features end-to-end.
What we’re looking for
- 4+ years building production backend systems.
- Strong in one of Go/Node/Python/Java; solid API design and testing.
- Datastores: Postgres/MySQL, Redis; messaging/queues (Kafka/SQS/NATS).
- Cloud-native ops (Docker/K8s, CI/CD), performance profiling, and security basics (OWASP, authN/Z).
- Product mindset: ship incrementally, measure, and iterate.
Nice to have
- Experience with search/retrieval (Elastic/OpenSearch), vector DBs, or RAG gateways.
- Payments/balance systems and idempotent workflows.
- Real-time features (WebSocket/Server-Sent Events).
Our stack (evolving)
Go/Node/Python, gRPC/REST, Postgres, Redis, Kafka, OpenSearch/Vector DB, Docker/K8s, Terraform, GitHub Actions, GCP/AWS.
Success in 90 days
- Stable, observable APIs for Q&A and feedback live with agreed SLOs.
- Clean service boundaries, rate-limits, and a clear runbook on-call.
Working at Andarz
AI-first, business-driven, user-obsessed. Competitive comp, meaningful equity, flexible hours, high ownership, and a product that improves how students learn. We’re an equal-opportunity team.