Staff Software Engineer (AI-Native)

Job description

## 𝗪𝗲’𝗿𝗲 𝗵𝗶𝗿𝗶𝗻𝗴 𝗿𝗼𝗰𝗸𝘀𝘁𝗮𝗿𝘀. We are looking for **Staff Software Engineers** who have stopped measuring their impact by how much code they personally write and started measuring it by how much better the entire engineering system performs. Gradion is hiring a **Staff Software Engineer, AI-Native** to work across systems, teams and technical domains. Your scope is bigger than one product or one technical layer. You will shape systems spanning web, desktop and mobile front ends, back-end services and APIs, data pipelines, infrastructure, testing, security and operations. You are a **software engineer first**. AI is how you work. The working rule on this team is **𝘇𝗲𝗿𝗼 𝗵𝗮𝗻𝗱-𝘄𝗿𝗶𝘁𝘁𝗲𝗻 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗰𝗼𝗱𝗲**. You work through Claude Code, Codex and whatever replaces them next, and you own every line that lands even though you do not type it. Your job is to turn outcomes into specifications, give agents the context and constraints they need, direct the implementation, and prove the result before it ships. At Staff level, that leverage extends beyond your own work. You shape how other engineers work with agents, establish technical standards, identify where the real difficulty sits, and create patterns that help teams move faster without lowering the engineering bar. **The implementation may be generated. The engineering judgement is yours.** You decide the boundaries, data models, failure behaviour, trade-offs, security controls, migration paths and operational concerns that AI will not decide for you. You know when to build, when not to build, and when a plausible answer from an agent is still the wrong answer. If your best work still comes from writing the hardest code yourself, this is probably not the right role. If you want your technical judgement to influence systems, teams and the way software gets built, keep reading. ## 𝗪𝗵𝗮𝘁 𝘁𝗵𝗶𝘀 𝗿𝗼𝗹𝗲 𝗶𝘀 𝗿𝗲𝗮𝗹𝗹𝘆 𝗮𝗯𝗼𝘂𝘁 - **Setting technical direction across systems and teams.** Turn ambiguous problems into architecture, constraints and an approach that engineers can execute against. - **Owning architecture and delivery for complex systems.** Stay involved from specification and design through implementation, testing, deployment and production operation. - **Designing how engineers work with AI agents.** Establish the conventions, repository context, tooling, evaluation methods and quality gates that make agents reliable on real production work. - **Making the decisions AI will not make for you.** Shape boundaries, data models, failure behaviour, scalability, security, migration paths and the trade-offs that are expensive to reverse. - **Raising the quality of engineering output.** Review the implementations carrying the most risk, set standards others can review against, and build mechanisms that catch bad output before it reaches production. - **Working across the full engineering surface.** Move comfortably between application architecture, APIs, front ends, infrastructure, CI/CD, testing and production issues when the problem demands it. - **Removing recurring problems at the source.** When several teams keep hitting the same issue, look beyond the immediate fix and change the architecture, platform, tooling or engineering practice behind it. - **Leading technical conversations with clients and stakeholders.** Take an unclear business problem, challenge assumptions, identify the important constraints and turn the conversation into a technical direction people can act on. - **Multiplying other engineers.** Raise the technical bar through reviews, architecture discussions, written guidance, pairing and better engineering practices rather than relying on formal authority. - **Creating reusable leverage.** Turn successful solutions into patterns, tooling, documentation or practices that other teams can adopt instead of solving the same problem repeatedly. ## 𝗪𝗵𝗮𝘁 𝘄𝗲’𝗿𝗲 𝗹𝗼𝗼𝗸𝗶𝗻𝗴 𝗳𝗼𝗿 - Around **nine years or more** building and running production software, including systems you have supported well beyond their initial release. - Strong software engineering fundamentals with **daily use of AI coding tools** such as Claude Code or Codex on real production work. You understand what to delegate, how to provide the right context and constraints, and how to verify what comes back. - Comfort with **zero hand-written production code** as the normal way of working. Your leverage comes from specification, architecture, context, review, evaluation and verification rather than manually typing implementation code. - Strong system-design capability across **scalability, performance, reliability, security and operational resilience**, with the judgement to identify failure modes that only become visible in production. - Breadth across the engineering stack: **web, desktop and mobile front ends, back-end services and APIs, CI/CD, cloud infrastructure, testing strategy and security**, with deep expertise in at least one area. - No specific stack required. You should be able to understand an unfamiliar codebase, architecture or technology quickly and make useful technical decisions without a predefined playbook. Our current projects use TypeScript, React, Flutter, Python, Go, Java, Node.js, PostgreSQL, AWS, GCP, Kubernetes and Terraform, among others. - A track record of **technical leadership beyond your own codebase**, whether across several teams, a large programme, a platform or a technically complex product. - English strong enough to lead technical discussions with clients, challenge requirements, write architectural direction and communicate decisions clearly enough for teams to execute without you in the room. - A strong **product and commercial mindset**. You understand that engineering exists to solve a business problem, and you weigh technical decisions against value, risk, time and what the client is actually paying for. - Experience across multiple business domains, including at least one where mistakes are expensive, such as **fintech, health tech, logistics or enterprise systems**. - Strong judgement about **when to build, when not to build, when to pay down technical debt, and when an agent has confidently gone in the wrong direction**. - The ability to influence without relying on hierarchy. People trust your technical direction because the reasoning is clear, the trade-offs are understood, and the result holds up. ## **𝗡𝗶𝗰𝗲 𝘁𝗼 𝗵𝗮𝘃𝗲** - Experience building and operating **production AI systems at scale**, including evaluation harnesses, guardrails, observability, regression testing, and cost or latency optimisation. - Experience building **AI agents as a product**, including tool calling, planning, RAG and human-in-the-loop design. - Experience defining or evolving **AI-native engineering practices** across a team or organisation. - Platform or internal tooling work that measurably improved the productivity, quality or reliability of other engineering teams. - Depth in **data platforms, cybersecurity, robotics or enterprise systems**. - Open source contributions, conference talks, technical writing or other evidence of sharing technical knowledge beyond your immediate team. ## 𝗪𝗵𝘆 𝗚𝗿𝗮𝗱𝗶𝗼𝗻 This is not a Staff role where you spend your time protecting one architecture diagram or reviewing pull requests all day. Gradion works across **AI, data, cybersecurity, robotics and enterprise platforms**, so the problems change and the engineering surface stays broad. You will work across teams and domains, influence how difficult technical problems are approached, and help define what modern software engineering looks like as AI changes the way software is built. We have been engineering for **23+ years**, with **300+ specialists across 7 countries** and more than **100 enterprise clients**. You will work with strong engineers on global projects, stay close to technology, and have real room to shape both the systems we build and the way we build them. **If you want Staff-level scope without leaving engineering behind, we should talk.** ## 𝗪𝗵𝘆 𝘆𝗼𝘂’𝗹𝗹 𝗹𝗼𝘃𝗲 𝘄𝗼𝗿𝗸𝗶𝗻𝗴 𝗵𝗲𝗿𝗲 - You’ll work in a global engineering environment where **AI and emerging technologies are part of real delivery**, not side projects or demos. We keep technical autonomy high and unnecessary bureaucracy low, so good engineering ideas can move. - You’ll have the scope to influence **architecture, engineering practices and AI-native ways of working** across teams rather than being limited to one product or technology stack. - Gradion has been recognised by **ITviec for 8 consecutive years, including Winner for the past 2 years**. You’ll collaborate across countries, work on technically challenging projects, and have access to continuous learning and tech communities. - We also offer competitive compensation, **up to 2 months’ bonus, twice-yearly performance reviews, premium healthcare, 15 days annual leave, full salary during probation**, and a hybrid working model. - **Working time:** Monday–Friday, 09:00–18:00 - **Locations:** Ho Chi Minh City, Da Nang, Hanoi and Bangkok