Staff or Senior Software Engineer, AI-Native SDLC

HubSync•,

10d

About The Position

HubSync is building an AI-native software development lifecycle (SDLC) to enhance its platform for top CPA firms. This role will own the entire lifecycle, from planning and writing to reviewing, testing, shipping, and operating software using AI agents. The goal is to treat internal agents as products, with AI reviewers, QA agents, ticket sharpeners, bug fixers, documentation writers, and production monitoring agents already in development. The successful candidate will be responsible for embedding with teams, building credibility, contributing to backlogs, and removing blockers to make this AI-native SDLC a reality.

Requirements

Demonstrable, unusual productivity with AI coding tools.
Experience living through an AI-native SDLC transition in a previous role.
Proficiency in running multiple agents in parallel.
In-depth knowledge of harnesses and context management for AI agents.
Ability to provide evidence (PRs, tools, automations) of recent AI tool usage.
Strong engineering fundamentals, including CI/CD, testing strategy, GitHub platform internals, and basic AWS infrastructure knowledge.
Ability to ship a working V1 in a day and iterate on it.
High degree of satisfaction from improving team productivity over individual speed.
Patience and credibility to win over skeptics with working software.
Comfortable being measured on adoption and dollars saved.
Ability to operate with a 'show, don't tell' approach.

Responsibilities

Build internal agents as products across the entire SDLC, including planning, writing, reviewing, testing, shipping, and operating.
Develop a shared harness layer for agents, including skills, context, MCP servers, and per-repo configuration.
Implement AI code reviewers, codified review rules, and measurement systems to earn merge authority.
Create PR-triggered QA agents, exploratory bug hunters, and generated Playwright suites.
Develop CI/CD gates, deploy verification, incident triage, and RCA drafting.
Embed with teams, lead by example by working on their backlogs, and showcase the capabilities of AI-native development.
Remove adoption friction related to permissions, infrastructure, environments, harness configuration, and prompt/context engineering.
Own the measurement of key metrics such as cycle time, review latency, escaped defects, cost per feature, and developer experience.
Maintain harnesses and playbooks, wire gates for code review and testing, and refine the doctrine based on field learnings.
Determine when an AI agent is the appropriate tool for a given task.