Agentic Delivery Supervisor

About the Role First Student is North America’s leading provider of student transportation, helping millions of students get to and from school safely each day. Our technology teams build and support the operational, safety, customer, and employee-facing systems behind that work, including our HALO platform. AI is an area of active investment for us, and this role is one of the ways we are building AI capabilities the business can rely on and govern. You will own First Student’s path to production for agent-built software. The role has no direct reports. You decide how code moves from an AI Software Supervisor’s environment into production, supervise the agents to the automated route that carries it there, and own how it behaves once it is live. This role is one seat in a three-person pod we are building toward. The other two are a Product Builder, who decides what to build and why, and an Agentic Software Supervisor, who directs coding agents to build it. Your part is making sure what they produce reaches production fast, holds up once it is there, and never puts student data or daily operations at risk. Depending on when you join, some of these seats may be filling in around you, and you will have real say in how the pod works and how the path to production gets built. Your results show up in two numbers: how fast code reaches production, and how safely. How You’ll Work You will use AI agents and tools to build and run the delivery path itself, not only to ship the product code that travels down it. Deployment automation, test and evaluation harnesses, policy checks, and monitoring are things you assemble and keep improving, the same way the Agentic Software Supervisor works with code generation. Your job is to define what counts as production‑ready here, prove it automatically, and decide where a human still has to sign off. A central part of the role is building the paved road: the golden paths, guardrails, and self‑service steps that let the pod ship without re‑deciding how deployment, testing, security, and rollback work every time. The goal is a route where the safe way to production is also the fast way, so no one has a reason to go around it. A typical week may include: Designing and building the automated path that carries agent-generated code from merge to production, with the gates and checks that keep it safe at agent speed. Standing up or extending CI/CD, infrastructure‑as‑code, and deployment automation on AWS. Building automated quality, security, and policy gates that catch defects in AI-generated code before they reach production. Operating deployment, test, and monitoring agents, and reviewing their output, so the path runs with little manual effort. Instrumenting production for observability, and owning rollback, incident response, and recovery when something breaks. Working with the Product Builder and Agentic Software Supervisor to close the gap between code being written and code being live. Key Responsibilities Own the path to production – Work closely with Dev manager & Cyber leader, decide and document how code moves from an AI Software Supervisor’s environment into production, and own that route end to end. Build and maintain CI/CD pipelines, infrastructure‑as‑code, and deployment automation that ship reliably on AWS. Make deployment fast and repeatable through progressive delivery (canary, blue/green, feature flags) so changes ship in small, reversible steps. Keep the path ahead of the rate coding agents generate code, clearing manual bottlenecks as volume grows. Own release readiness and what actually ships, beyond the mechanics of the pipeline. Build and operate the agentic delivery platform – build the golden paths, scaffolds, and self‑service workflows that let the pod deploy without re‑solving deployment each time. Encode standards into the platform as reusable defaults and automated checks, so the right way is built in instead of written down and forgotten. Build for more than today’s pod, so the path can support additional builders and teams as agentic delivery spreads across IT. Build automated quality gates (unit, integration, end‑to‑end, contract, and evaluation checks) that verify AI‑generated code at machine speed. Encode security, privacy, and compliance into the path as policy‑as‑code and safe defaults, so the obligations attached to student and driver data, including FERPA, hold by construction. Manage secrets, access, and least‑privilege provisioning across environments, including for the agents operating inside the pipeline. Instrument production with logging, metrics, and monitoring so problems surface early and are easy to diagnose. Own rollback, incident response, and recovery, and drive down both how often changes fail and how long recovery takes. Partner with architecture, security, and AI governance to keep the path aligned with enterprise standards. Required Qualifications 5+ years building and operating production delivery systems (CI/CD, infrastructure‑as‑code, deployment automation, or platform/SRE work), with senior‑level judgment about what is safe to ship. A track record of shipping software to production reliably and repeatably, and owning what happens after deploy. Strong infrastructure‑as‑code and cloud deployment skills, ideally on AWS (for example Terraform or CDK, containers, and managed compute). Strong automated testing and quality‑gate practices, including using tests and checks to validate code you did not write by hand. Practical experience with observability, monitoring, incident response, and rollback in production. Experience building security and compliance into delivery: secrets management, least privilege, and policy or guardrail enforcement in pipelines. Demonstrated experience using AI coding agents or AI‑assisted tools to ship real work, beyond autocomplete or experimentation. Ability to break a delivery process into clear, automatable steps and build the automation that runs it. Experience in enterprise environments with security, governance, data‑handling, and production‑support constraints. Strong written and verbal communication, including explaining delivery tradeoffs and risk to technical and non‑technical stakeholders. Bachelor’s degree in Computer Science, Engineering, a related field, or equivalent practical experience. Preferred Qualifications Direct experience deploying React and React Native applications on AWS. Experience with AWS delivery and AI services such as CodePipeline, CodeBuild, ECS/Fargate, Lambda, API Gateway, Step Functions, S3, DynamoDB, RDS, or Bedrock. Experience building internal developer platforms, golden paths, or paved‑road workflows (for example Backstage or equivalent). Experience with progressive delivery (canary, blue/green, feature flags) and supply‑chain security (SBOMs, artifact signing, dependency scanning). Experience with policy‑as‑code (for example OPA) and compliance‑by‑construction in regulated environments. Experience building evals, guardrails, monitoring, or quality gates for AI‑generated or AI‑embedded systems. Experience with advanced testing such as contract, property‑based, or mutation testing. Experience with geospatial, routing, logistics, transportation, or operations‑focused software. Experience applying AI within governed enterprise environments, including FERPA‑relevant or similarly regulated data. Relevant AWS, security, SRE, or platform certifications. What Success Looks Like At 6 months – a working path to production that the pod uses to ship without hand‑built deployment each time; automated quality, security, and policy gates that catch common defects in agent-generated code; baseline observability, rollback, and incident response in place; and the path trusted enough to be used routinely. At 12 months – the path keeps pace with the pod’s output, and both lead time to production and change‑failure rate have measurably improved; the delivery platform supports more than the original pod and sets the pattern for how agent-built software ships across IT; golden paths, automated gates, and observability are reused by other teams. Working Style The role takes initiative and judgment. You will define a path where none exists, make progress with limited direction, raise risk early, and stay accountable for the quality and behavior of what ships. Speed here depends on discipline. You will set clear gates, prove safety automatically, and keep ownership of production even when agents wrote the code moving through it. Equal Employment Opportunity Statement All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability or veteran status. First is also committed to providing a drug‑free workplace. First will consider for employment qualified applicants with criminal histories consistent with the requirements of the San Francisco Fair Chance Ordinance, Los Angeles Fair Chance Ordinance, and any other fair chance law. Philadelphia’s Fair Criminal Record Screening Standards Ordinance poster is available upon request. #J-18808-Ljbffr FIRST MANAGEMENT SERVICES LLC