We are an AI-first company building innovative agentic frameworks, the systems that let LLM-driven agents research, plan, and act reliably in production. That means orchestrating multi-step workflows with retries, skips, and parallel execution; keeping long-running work persistent and observable; and serving per-tenant retrieval at scale. This isn't about wrapping an LLM in an API call, it's the hard backend engineering that makes agents dependable.
Our stack is Python (FastAPI) on AWS, with Terraform. We hire on engineering depth and judgment, not on a keyword count, and we work in whichever language you're strongest in during the interview (more below). As a Senior Engineer, you will own backend services and the orchestration layer on top of them, from design through deployment, operation, and the on-call that follows.
Our team is based in Tehran and pays salaries monthly in USD.
What You Will Do:
Design, build, and own backend services and the agent-orchestration layer end to end.
Build resilient systems that stay correct under load and degrade gracefully when a model, tool, or dependency fails.
Ensure operational excellence by making long-running, multi-step work observable and debuggable in production.
Shape technical direction through rigorous code reviews, mentorship, and pushing back on specifications when they misalign with engineering reality.
Drive the full software development lifecycle, translating ambiguous requirements into durable architectures and actionable technical plans.
What We Are Looking For:
5+ years building and operating production backend systems, with deep ownership of at least one non-trivial service you can talk through end to end.
Deep fluency in your primary backend language, and the ability to pick up a new one fast. We work in Python; if it is not your daily language today, we expect you to ramp quickly.
Strong systems fundamentals — concurrency, data modeling, and failure modes — and the judgment to choose the right abstraction rather than the cleverest one.
Sound reasoning under ambiguity: you ask the clarifying question first, make the trade-offs explicit, and adapt when the constraints change.
Solid grasp of infrastructure and operations, including message brokers (Kafka, RabbitMQ), SQL and NoSQL stores, and running services in containers with an orchestrator (Kubernetes, ECS, Nomad, or similar).
Professional written English, enabling you to write crisp design documents and align stakeholders on complex technical trade-offs.
Even Better:
AWS (especially ECS) and Terraform, or comparable cloud and Infrastructure-as-Code experience.
Hands-on work with LLM, agentic, or RAG systems (LangChain, CrewAI, OpenAI/Anthropic SDKs, Hugging Face) and a real grasp of how non-deterministic agent steps differ from ordinary functions.
A recognized-bar engineering background, owned open-source work, or another verifiable standout signal.
How We Hire:
We use a deliberate, transparent two-stage process designed to see how you think and build:
1. A 120-minute conversation focusing on your background and a collaborative system-design exercise and cultural fit..
2. A 120-minute live technical session involving code review, live coding, and a self-review.
You may complete the technical session in your strongest language (Python, Go, Java, Kotlin, PHP, or TypeScript/Node.js). Web search is encouraged; AI assistants are not used. We want to see your fundamental engineering judgment.
What We Offer:
Competitive USD-based salary and a comprehensive benefits package.
Hybrid work flexibility with hours that support deep work.
The opportunity to work at the frontier of agentic AI, tackling hard problems with high ownership.
A highly educated, agile senior team that values pragmatism, learning, and continuous improvement.
If you would rather be judged on how you think than on a buzzword list, we would like to talk!
Originally posted on