Mô tả công việc
The kinds of problems you'll solve
❖ Taking LLM-based systems (RAG pipelines, multi-agent workflows, conversational assistants) from "works in a demo" to reliable under real-world, messy, adversarial, or incomplete input.
❖ Building the evaluation and quality layers most AI teams lack: regression testing for prompts/agents, hallucination and safety checks, retrieval quality measurement, latency/cost/quality tradeoffs.
❖ Adding the operational maturity that's usually missing from AI prototypes: observability, alerting, retries/fallbacks, security review, and incident-readiness.
❖ Improving cloud infrastructure and deployment pipelines so services scale, restart cleanly, and cost less to run.
❖ Deciding, system by system, what "production-ready" actually requires - and pushing back when a shortcut isn't safe to ship.
❖ Mentoring other engineers and raising the bar on testing, code quality, and documentation across the team.
Tech stack you'll work in
❖ AI/LLM engineering: LLM application development, RAG, agentic/tool-using systems, prompt engineering, evaluation and guardrails.
❖ Backend: Python (modern tooling - package managers like uv, linters like Ruff), with some Node/TypeScript.
❖ Frontend: React-based frameworks ([protected info]), with internationalization considerations.
❖ Cloud & infrastructure: AWS (containers, serverless functions, CDN/static hosting), infrastructure-as-code, Docker.
❖ Quality & delivery: End-to-end testing (Playwright or similar), backend test suites, CI/CD pipelines.
Yêu cầu
❖ 6+ years in
software engineering, including 2+ years shipping LLM/AI-powered applications to production - not just prototypes or notebooks.
❖ Strong, practical experience with RAG and/or agentic AI systems and their real failure modes (bad retrieval, hallucination, tool-call errors, context limits).
❖ Solid Python; comfortable reading/writing in a JS/TS frontend codebase.
❖ Experience owning services on a major cloud provider with infrastructure-as-code and containerized deployments.
❖ Demonstrated ability to make AI systems reliable and measurable - testing, evaluation, monitoring, and cost/performance tuning.
❖ Good product judgment - able to scope "production-ready" pragmatically rather than gold-plating or cutting corners.
❖ Good English communication
Nice to have
❖ Experience with computer vision, structured-data/SQL generation, or real-time data pipelines.
❖ LLMOps tooling (evaluation frameworks, tracing/observability platforms).
❖ Security or compliance-sensitive project experience (finance, government, or regulated industries).
❖ Experience with multilingual or RTL-language products.
Quyền lợi
At SmartOSC, we offer the best to your values:
❖ You'll take real AI systems the last, hardest mile to production - with the autonomy of a senior engineer and real influence over how the team builds AI going forward.
❖ Attractive salary package
❖ Salary review twice a year
❖ Flexible working hours (between 7:30am - 7:30pm on staff's preference)
❖ Premium health care
❖ Working in one of the largest digital transformation agencies in Asia - Professional English environment
❖ Firm's Certified Qualifications Sponsorship for career development
❖ Annual company trip inside or outside Vietnam
❖ Other fun activities: happy hour, quarterly team building, football club, yoga club, swimming club, charity activities, etc.
❖ Free entertainment parties: Birthday party, Anniversary party, Sum-up Party, Year-End Party, etc
Thông tin khác
Thời gian làm việc
Thứ 2 - Thứ 6 (từ 08:00 đến 17:00)
Thông tin chung
Nơi làm việc
- - Hà Nội: Phường Cầu Giấy (quận Cầu Giấy cũ)