About STYL & the Agentic Delivery Model
STYL has moved software delivery away from line-by-line manual coding to an AI-centric Agentic Delivery Model. Our squads do not simply write code - they command industrialised production lines of multi-skilled AI coding agents, synchronised over an enterprise event-driven architecture. Because frontier models and coding harnesses improve week by week, our delivery methods adapt continuously.
We are seeking an AI Fleet Commander to direct and optimise these parallel production lines within our Ho Chi Minh City delivery team, adapting fluently to STYL's agentic delivery approach and its OpenSpec pipeline under a governed human sign-off.
Key Responsibilities
1. Command & orchestration of the agent fleet
Direct a parallel fleet of multi-skilled AI agents - scoping work into verifiable units and delegating it through STYL's delivery pipeline.
Manage context, prompt chaining and loop dependencies so agents produce clean, decoupled service logic.
Collaborate with engineers and
product owners on the delivery team to keep agent output aligned with business context.
2. Verification & quality assurance
Review agent work across front-end, back-end, data and infrastructure; accept or reject each change with a stated reason - never rubber-stamping, never rejecting wholesale.
Anticipate how a task is likely to fail and catch the serious, latent defects that pass a demo and the tests.
3. Specification-driven delivery (OpenSpec)
Own STYL's specification-driven approach: author and refine strict, machine-readable blueprints (API / AsyncAPI schemas and agent manifests) that serve as the ground truth for the fleet.
Operate the OpenSpec pipeline end-to-end and maintain a clear audit trail.
4. Governance, escalation & production-readiness
Escalate, gate or proceed under STYL's governance model (Human Governance Gate, Watchdog, QPA).
Distinguish production-ready from demo-ready - CI/CD, observability, scaling - not merely correct in isolation.
Enforce clean Git workflow: review agent-generated branches, pull-requests and merges so parallel lines integrate without regression.
5. Fleet economics & continuous adaptation
Understand and manage the cost and throughput of the agent sessions under supervision.
Rapidly evaluate and adopt emerging tools, coding harnesses and model upgrades, and build custom skills or tools to unblock the production line.
Core Competencies & Experience Required
Educational foundation - Bachelor's degree in Computer Science,
Software Engineering or a related technical discipline.
Engineering fundamentals & full-stack breadth - strong, language-agnostic software-engineering fundamentals - design, data, security, performance and testing - that hold in any stack; able to read and evaluate code across front-end, back-end, data and infrastructure, deep in at least one area and competent across the rest.
Data & systems reasoning - data modelling, transactions and concurrency, and the architecture and scaling trade-offs behind a design.
A supervisor's mindset toward AI - hands-on with AI-assisted coding tools and multi-agent workflows, and healthily skeptical of their output; able to explain why a change is accepted or rejected and name the trade-off it makes.
Collaboration & communication - strong communication and the ability to operate effectively within a collaborative engineering team.
AWS Associate certification - required to apply; candidates must hold a current AWS Associate-level certification - either Solutions Architect - Associate (SAA-C03) or
Developer - Associate (DVA-C02) - at the time of application