Senior Full Stack Developer, Agentic Engineering

Overview

This role is hired to build agent-first. The Senior Full-Stack Developer, Agentic Engineer is a founding member of a small, new engineering squad building the next generation of staff-facing applications, APIs, and data services for Catapult Learning and SESI Schools, the tools teachers, supervisors, and operations staff use every day. The squad starts with net-new functionality and then rebuilds existing capabilities one at a time. It is not responsible for maintaining the current legacy systems. AI coding agents are the default way this engineer plans, implements, tests, and refactors, and they own features end to end, from requirements and technical design through prototype, production code, and deployment.

Just as important, they help shape how the new squad works. FullBloom is moving its engineering teams from ad hoc AI-assisted coding to a shared agentic build pipeline. That pipeline runs on agent-ready specs, AI-driven implementation and testing, automated quality gates, and human review of everything that ships. This engineer helps stand that pipeline up, codifies what works, and mentors the early-career engineers who join as the squad grows. Because agents make code cheap to write, the hardest parts of this job are deciding what to build, defining it clearly, and judging whether what the agents produced is right.

They work with product, design, QA, DevOps, and security partners, often directly from lightweight product briefs. They move faster without compromising student data protection, accessibility, or SOC 2 controls.

Responsibilities

Agentic engineering and delivery

Agent-first development: Use AI coding agents (e.g., Claude Code) as theprimary way to plan, implement, test, and refactor code, while stayingaccountable for every line that ships

Judging agent output: Treat review and validation as the core of the job:catch defects in code that looks correct, decide which changes need closerhuman review, and own what merges

Requirements to architecture: Work directly with product partners andstakeholders to turn lightweight briefs into agent-ready specs, technicaldesigns, data flows, and acceptance criteria

Proof-of-concept development: Prototype quickly to test feasibility, showvalue, and settle design questions before committing to a full build

Short, experimental cycles: Build end-to-end working skeletons first (Simple,Lovable, Complete), apply YAGNI to keep AI-generated scope in check, andthrow away work that doesn't earn its place

Full-stack ownership: Own features end to end across front-end interfaces,APIs, back-end services, data stores, infrastructure as code, and deploymentpipelines

Integrations and data services: Build and maintain integrations with identityproviders, rostering and student information systems, internal APIs, andshared data platforms

Rebuilding legacy capabilities: Use agents to read existing systems wellenough to extract their business rules, data, and edge cases, then rebuildthose capabilities one at a time on the new stack

Building the agentic pipeline

Context and standards: Write and maintain the artifacts agents work from,including repository instruction files (CLAUDE.md/AGENTS.md), AI-readable coding standards, architecture decision records, and a shared skills andprompt library

Tooling and integrations: Configure agent tooling, permissions, and MCPintegrations (e.g., GitHub, Jira, browser automation) so agents can work safelyin our repositories and environments

Environments and DevOps: Own the squad's DevOps alongside the code,including preview environments, safe test data with student informationremoved, and deployment pipelines, so agents can build and verify changeswithout touching production data

Quality gates: Strengthen automated tests, CI checks, and AI-assisted codereview so agent output meets the same bar as human code before it merge

Verification loops: Build feedback loops that let agents check their own work,such as end-to-end tests and headless browser checks, while keepingindependently written tests for high-risk areas

Governance: Apply guardrails for secrets, data access, and student data, andkeep agent-driven work aligned with SOC 2 change-management and SDLC controls

Growing the squad

Mentoring: Mentor early-career engineers who join the squad, with a focus onthe fundamentals and review habits that agents don't teach

Codifying practice: Turn what works into repeatable playbooks, templates,and skills that new squad members and FullBloom's other engineering teamscan reuse, and log friction points so the pipeline improves with each project

Working across roles: Work closely with product and design partners in amodel where roles blur, including designers submitting changes againstworking code

Measuring outcomes: Track delivery and quality signals such as lead time,rework, escaped defects, and change-failure rate, not tool adoption alone

Quality, reliability, and collaboration

Review human- and agent-written code for correctness, security,maintainability, and performance

Monitor, troubleshoot, and resolve production issues, and take part in incidentresponse and post-incident reviews

Communicate scope, timelines, and trade-offs clearly to technical and non-technical colleagues

Promote usability, accessibility, and security best practices

Other duties as assigned

Qualifications

Required Qualifications

Bachelor's degree in Computer Science or a related field and 5+ years of relevant work experience, or equivalent experience

Daily, hands-on use of AI coding agents (e.g., Claude Code, Codex, Cursor) to build production software, with concrete examples of how it changed your workflow and results

Working knowledge of what makes agents effective: context management, task decomposition, clear specs and prompts, and rigorous verification of agent output

Demonstrated judgment reviewing agent output: you can find and explain problems in code that looks correct, and you know when a change needs closer human review

Strong product sense: you can turn an ambiguous problem into a clear, well-scoped spec and push back on features that aren't needed

Strong proficiency in TypeScript/JavaScript, React, HTML, and CSS

Experience designing and building REST APIs and back-end services

Strong working knowledge of relational databases (e.g., PostgreSQL), including data modeling, query writing, and performance tuning

Experience with cloud hosting, serverless technologies, and infrastructure as code (e.g., AWS with Terraform or CDK)

Proficiency in writing and automating unit, integration, and end-to-end tests

Experience with CI/CD pipelines such as GitHub Actions, and owning your own environments and deployments

Git and version control expertise, including pull-request-based review workflows.

Strong knowledge of cross-browser, cross-platform, and cross-devicecompatibility

Experience building accessible applications that meet WCAG and ADA requirements

A track record of owning features end to end and mentoring other engineers

Clear written communication, including specs and technical designs that both people and agents can work from

Comfort working in short, experimental cycles and discarding work thatdoesn't pan out

Preferred qualifications

Experience leading or contributing to a team's adoption of agentic development practices

Experience building agent tooling, such as MCP servers, custom skills orsubagents, or CI-integrated agent workflows (e.g., ticket-to-pull-request automation)

Experience building LLM-powered or multi-agent features with frameworks such as the Claude Agent SDK, LangGraph, or Google ADK, including tooldesign, orchestration, and memory

Experience rebuilding legacy applications on modern stacks, including extracting business rules from systems you didn't build

Experience as an early member of a new team or product

Experience designing data-heavy applications that handle large datasets with low latency and high reliability

Experience building education technology, especially for PreK-12 schools and districts

Experience integrating SSO and identity providers, including SAML or OIDC

Experience with education data integrations such as SIS rostering (e.g.,OneRoster, Clever) or LMS platforms

Experience building search or analytics features with Elasticsearch or similar engines

Familiarity with UX and HCI principles and design tools such as Figma

Experience working under student-data privacy requirements (FERPA, COPPA)or SOC 2 controls

Posted Salary Range

USD $120,000.00 - USD $150,000.00 /Yr. Physical Requirements

Remote work setting

Agentic engineering and delivery

Agent-first development: Use AI coding agents (e.g., Claude Code) as theprimary way to plan, implement, test, and refactor code, while stayingaccountable for every line that ships

Judging agent output: Treat review and validation as the core of the job:catch defects in code that looks correct, decide which changes need closerhuman review, and own what merges

Requirements to architecture: Work directly with product partners andstakeholders to turn lightweight briefs into agent-ready specs, technicaldesigns, data flows, and acceptance criteria

Proof-of-concept development: Prototype quickly to test feasibility, showvalue, and settle design questions before committing to a full build

Short, experimental cycles: Build end-to-end working skeletons first (Simple,Lovable, Complete), apply YAGNI to keep AI-generated scope in check, andthrow away work that doesn't earn its place

Full-stack ownership: Own features end to end across front-end interfaces,APIs, back-end services, data stores, infrastructure as code, and deploymentpipelines

Integrations and data services: Build and maintain integrations with identityproviders, rostering and student information systems, internal APIs, andshared data platforms

Rebuilding legacy capabilities: Use agents to read existing systems wellenough to extract their business rules, data, and edge cases, then rebuildthose capabilities one at a time on the new stack

Building the agentic pipeline

Context and standards: Write and maintain the artifacts agents work from,including repository instruction files (CLAUDE.md/AGENTS.md), AI-readable coding standards, architecture decision records, and a shared skills andprompt library

Tooling and integrations: Configure agent tooling, permissions, and MCPintegrations (e.g., GitHub, Jira, browser automation) so agents can work safelyin our repositories and environments

Environments and DevOps: Own the squad's DevOps alongside the code,including preview environments, safe test data with student informationremoved, and deployment pipelines, so agents can build and verify changeswithout touching production data

Quality gates: Strengthen automated tests, CI checks, and AI-assisted codereview so agent output meets the same bar as human code before it merge

Verification loops: Build feedback loops that let agents check their own work,such as end-to-end tests and headless browser checks, while keepingindependently written tests for high-risk areas

Governance: Apply guardrails for secrets, data access, and student data, andkeep agent-driven work aligned with SOC 2 change-management and SDLC controls

Growing the squad

Mentoring: Mentor early-career engineers who join the squad, with a focus onthe fundamentals and review habits that agents don't teach

Codifying practice: Turn what works into repeatable playbooks, templates,and skills that new squad members and FullBloom's other engineering teamscan reuse, and log friction points so the pipeline improves with each project

Working across roles: Work closely with product and design partners in amodel where roles blur, including designers submitting changes againstworking code

Measuring outcomes: Track delivery and quality signals such as lead time,rework, escaped defects, and change-failure rate, not tool adoption alone

Quality, reliability, and collaboration

Review human- and agent-written code for correctness, security,maintainability, and performance

Monitor, troubleshoot, and resolve production issues, and take part in incidentresponse and post-incident reviews

Communicate scope, timelines, and trade-offs clearly to technical and non-technical colleagues

Promote usability, accessibility, and security best practices

Other duties as assigned

Required Qualifications

Bachelor's degree in Computer Science or a related field and 5+ years of relevant work experience, or equivalent experience

Daily, hands-on use of AI coding agents (e.g., Claude Code, Codex, Cursor) to build production software, with concrete examples of how it changed your workflow and results

Working knowledge of what makes agents effective: context management, task decomposition, clear specs and prompts, and rigorous verification of agent output

Demonstrated judgment reviewing agent output: you can find and explain problems in code that looks correct, and you know when a change needs closer human review

Strong product sense: you can turn an ambiguous problem into a clear, well-scoped spec and push back on features that aren't needed

Strong proficiency in TypeScript/JavaScript, React, HTML, and CSS

Experience designing and building REST APIs and back-end services

Strong working knowledge of relational databases (e.g., PostgreSQL), including data modeling, query writing, and performance tuning

Experience with cloud hosting, serverless technologies, and infrastructure as code (e.g., AWS with Terraform or CDK)

Proficiency in writing and automating unit, integration, and end-to-end tests

Experience with CI/CD pipelines such as GitHub Actions, and owning your own environments and deployments

Git and version control expertise, including pull-request-based review workflows.

Strong knowledge of cross-browser, cross-platform, and cross-devicecompatibility

Experience building accessible applications that meet WCAG and ADA requirements

A track record of owning features end to end and mentoring other engineers

Clear written communication, including specs and technical designs that both people and agents can work from

Comfort working in short, experimental cycles and discarding work thatdoesn't pan out

Preferred qualifications

Experience leading or contributing to a team's adoption of agentic development practices

Experience building agent tooling, such as MCP servers, custom skills orsubagents, or CI-integrated agent workflows (e.g., ticket-to-pull-request automation)

Experience building LLM-powered or multi-agent features with frameworks such as the Claude Agent SDK, LangGraph, or Google ADK, including tooldesign, orchestration, and memory

Experience rebuilding legacy applications on modern stacks, including extracting business rules from systems you didn't build

Experience as an early member of a new team or product

Experience designing data-heavy applications that handle large datasets with low latency and high reliability

Experience building education technology, especially for PreK-12 schools and districts

Experience integrating SSO and identity providers, including SAML or OIDC

Experience with education data integrations such as SIS rostering (e.g.,OneRoster, Clever) or LMS platforms

Experience building search or analytics features with Elasticsearch or similar engines

Familiarity with UX and HCI principles and design tools such as Figma

Experience working under student-data privacy requirements (FERPA, COPPA)or SOC 2 controls