Agentic Software Engineer – Full Stack

This is a fully remote position, open to applicants in Egypt.

📋 Description

• Design, develop, and implement agentic products from start to finish, including chat-based interfaces, AI workflow tools, and internal copilots.

• Utilize agentic coding tools like Claude Code, Codex, Cursor, and similar assistants while ensuring high code quality and adherence to engineering standards.

• Architect the internals of agents, covering tool invocation, multi-step orchestration, memory and state management, retrieval pipelines, permissions models, and strategies for failure recovery.

• Construct and manage agent evaluation infrastructure, incorporating regression suites, scenario-based tests, golden traces, and both offline and online metrics.

• Implement observability suitable for production, which includes distributed tracing, tracking of prompts and versions, telemetry for tool calls, and monitoring of costs/latency.

• Take ownership of the entire production lifecycle, which includes rollout strategies, quality assurance, guardrails, incident response, and the ongoing enhancement of deployed agents.

⛳️ Requirements

• Demonstrated full-stack engineering capabilities with production software delivered from end to end.

• Hands-on experience in building LLM-powered agents that are operational in production environments.

• Responsible for at least 1–2 agentic features or products from deployment through their operational phase.

• Profound understanding of agentic coding best practices, including task decomposition, agent delegation, prompt hygiene, iterative loops, and reviewing and guiding agentic outputs.

• Practical experience in agent evaluations and tracing, which encompasses trace-based debugging and regression detection.

• Strong engineering fundamentals related to testing, debugging, security principles, performance optimization, and reliability.

• Proficient in TypeScript/Node.js and/or Python (preferred).

• Experience with contemporary web stacks and the ability to deliver at a startup pace (preferred).

• Familiarity with prompt/version stores, evaluation harnesses, and tracing and telemetry platforms like OpenTelemetry, LangSmith, Datadog, and Honeycomb (preferred).

• Experience with RAG architectures, vector databases, indexing strategies, chunking methodologies, and retrieval evaluation (preferred).

• Experience in designing tool APIs for agents, including structured output schemas, idempotency, rate limiting, and error handling patterns (preferred).

🏝️ Benefits

• 100% remote work.

• Open and relatively flat organizational structure.

• Collaborative culture that emphasizes the sharing of ideas and information.

• Inclusive atmosphere with a strong commitment to diversity.

• Entrepreneurial, informal, adaptable, intelligent, and fast-paced work environment.