Phát triển hệ thống backend mở rộng và các quy trình AI (workflow, RAG, agent) cho nền tảng Everfit.
Yêu cầu
Từ 4 năm kinh nghiệm lập trình backend chuyên nghiệp
Kinh nghiệm lập trình vững với Node.js, TypeScript hoặc Python
Nắm vững REST API, microservices, hệ thống phân tán và cơ sở dữ liệu
Đã từng đưa tính năng LLM vào production qua API của OpenAI hoặc Anthropic
Giao tiếp tốt bằng tiếng Anh trong môi trường quốc tế
Trách nhiệm
Xây dựng workflow AI chạy lâu trên Temporal, chịu được restart và lỗi từng phần
Xây dựng pipeline AI có validation, retry, fallback, guardrails và observability
Đưa năng lực AI đã kiểm chứng từ đội AI thành service production
Xây dựng evals và monitoring để đo chất lượng AI, phát hiện regression sớm
Có thì tốt
Kinh nghiệm với Python, FastAPI, xây dựng pipeline RAG hoặc hệ thống agent
Quen thuộc với stack Pydantic, LangGraph, Qdrant, Langfuse, MCP
Kinh nghiệm với AWS, Docker, message queue hoặc kiến trúc cloud-native
Lương và phúc lợi
Thưởng gia nhập (Sign-on Bonus)
Tóm tắt bằng AI từ JD gốc, đọc JD gốc trước khi ứng tuyển.
About the Role
As an Applied AI Engineer at Everfit, you'll work at the intersection of backend engineering and applied AI, building the scalable systems and AI-powered experiences that shape the future of fitness coaching for coaches and their clients around the world.
This is not a Backend role with AI sprinkled on top. Expect roughly 70% of your work on LLM-powered systems (agent workflows, RAG, tool calling, evals) and 30% on the core backend services that power them.
Responsibilities
What you'll build
Durable, long-running AI workflows on Temporal: scheduled jobs and multi-step agent processes that survive restarts, retries, and partial failures.
Reliable AI pipelines with validation, retries, fallbacks, guardrails, and observability.
Systems designed around AI-specific constraints: latency, token cost, model failures, quality.
Production integrations with OpenAI and Anthropic models.
The path from experiment to production: take a validated AI capability from the AI team and ship it as a service with clear reliability, cost, and quality characteristics.
Evals and monitoring that define how AI features are measured, and catch regressions before users do.
How you'll work
Scalable backend services in Node.js/TypeScript, with system architecture, technical design, and code review across Everfit's backend.
Rapid experiments with emerging AI tech, turning the successful ones into production solutions.
Day-to-day with Product Managers, AI Engineers, and QA, from technical discovery through delivery.
Investigating production issues and improving AI workflows from real user feedback.
Helping the broader engineering org adopt AI-native development practices.
Requirements
4+ years of professional backend engineering experience.
Strong programming experience (preferably Node.js / TypeScript / Python) and backend architecture. Our core backend is Node.js/TypeScript and our AI services are Python/FastAPI; you'll work mainly in Node but should be comfortable crossing into Python when the work lives there.
Solid understanding of REST APIs, microservices, distributed systems, databases, and scalable system design.
Strong experience with databases such as MongoDB, PostgreSQL, and Redis.
You have shipped at least one LLM-powered feature to production for real users, working directly with OpenAI or Anthropic APIs, and can talk concretely about its failure modes, evals, latency, and cost.
You already use AI coding tools (Claude Code, Codex) daily, and you review what they produce with judgment about architecture, security, and edge cases.
Comfortable experimenting with new AI tools and technologies and turning successful experiments into production solutions.
Comfortable communicating in English and working with an international team.
Curious, proactive, and excited about how AI can fundamentally change software development and product experiences.
Preferred Skills
Experience with Python / FastAPI or AI-focused backend development.
Experience building production LLM features, RAG pipelines, or agentic systems.
Familiarity with the stack we run in production: Pydantic/LangGraph, Qdrant for vector search, Langfuse for LLM observability, and MCP.
Experience with AWS, Docker, message queues, or cloud-native architecture.
Experience improving engineering productivity through AI, with measurable impact.
Experience working with health, fitness, wellness, personalization, or recommendation systems.
Open-source contributions or personal projects involving AI.
