Founding Engineer: Hands-On, Agent-First

Full-time from day one · Remote worldwide · Competitive salary plus meaningful equity

About FrontierU

FrontierU is building a private global community for selectively admitted young people aged 16 and over. Members connect with ambitious peers, collaborate, learn, and receive personalised AI mentorship from Socrates.

A substantial working platform and automated test suite already exist. We are hiring a hands-on CTO to take full ownership of the technology, complete the product, launch it, and run it.

The role

This is not a traditional executive role for someone who has moved away from engineering.

You will personally direct the architecture, development, integration, deployment, security, and operation of FrontierU. You will use AI agents as your primary source of engineering capacity through the agentic harness—or combination of tools—you judge best.

No particular harness is required. You will have the authority to select, combine, replace, or build the tools that produce the best results.

You may build a human engineering team where people add expertise, judgment, or operational coverage. However, FrontierU intends to remain lean. You should know how to get exceptional output from agents before solving every capacity problem through hiring.

You will work directly with the founder and have broad authority over technical execution. You will also be accountable for everything your agents produce.

Your objectives

Within your first 30 days: achieve full prototype parity. Your first objective is to finish the platform so the implemented product matches FrontierU’s approved product prototype. The prototype, discrepancy analysis, backlog, and acceptance criteria are already documented. Significant parts of the product are already built.

You will:

- Take control of the codebase, architecture, backlog, development environments, documentation, and existing agent workflows.

- Turn the remaining work into clear, independently executable lanes for AI agents.

- Complete the remaining product experiences, backend behaviour, integrations, and prototype discrepancies.

- Integrate and review work that is complete or already in progress.

- Ensure every feature works end to end. A rendered interface without correct permissions, persistence, error handling, and tests does not count as complete.

- Preserve FrontierU’s access controls, safeguards for younger Members, audit requirements, and privacy guarantees.

- Keep the automated tests and all engineering quality gates green.

- Demonstrate full prototype parity on a production-like staging environment by the end of the month.

Any dependency requiring a founder, legal, infrastructure, or vendor decision must be identified during the first week, with a clear recommendation and decision deadline. It must not remain a hidden blocker.

Within your first 90 days: launch FrontierU

After reaching prototype parity, you will make the platform ready for real users and lead its production launch.

You will:

- Run structured founder testing, closed-beta testing, and end-to-end testing of all critical journeys.

- Resolve every launch-critical defect and establish a disciplined process for triaging everything else.

- Complete the production configuration of infrastructure and external services.

- Maintain the required European data-residency and privacy posture.

- Establish monitoring, structured logging, alerts, health checks, operational dashboards, and incident response.

- Put automated backups in place and prove that production data can be restored.

- Document deployment, database migration, rollback, recovery, and security procedures.

- Verify the security baseline, including authentication, authorization, rate limiting, secrets, staff permissions, auditability, encryption, and abuse prevention.

- Complete accessibility testing of public and Applicant-facing journeys and resolve material violations.

- Verify the platform against agreed performance targets using production-like data.

- Launch the platform and remain responsible for its operational stability.

You will:

- Maintain the production system and respond quickly when Members or Staff surface bugs.

- Own incident response, root-cause analysis, security updates, dependency upgrades, capacity, performance, and infrastructure costs.

- Define and meet appropriate reliability and response-time objectives.

- Deliver the next stages of the product roadmap.

- Improve FrontierU’s AI systems through evaluations, safety testing, model comparison, and measured iteration.

- Turn the agent workflow into a durable engineering operating system that others can use.

- Continuously evaluate new models, harnesses, protocols, and orchestration techniques.

- Adopt new tools when evidence shows they improve quality, safety, cost, or delivery.

- Avoid unnecessary dependence on one agent vendor or workflow.

- Hire and lead a small, high-quality human engineering team where justified.

- Maintain clear technical documentation so the platform is not dependent on one person’s memory.

- Help set product priorities and translate them into technically sound delivery plans.

What agent-first engineering means here

We are not looking for someone who occasionally asks an AI assistant to complete a function. We also do not prescribe a particular agentic harness. Selecting and evolving the toolchain is part of your job. You should have strong practical experience setting up and running agents to deliver production software. You should be comfortable with:

- Comparing agentic harnesses based on quality, control, cost, security, and fit.

- Breaking large objectives into bounded tasks with explicit acceptance criteria and dependencies.

- Running multiple agents across isolated branches, worktrees, sandboxes, or environments without creating integration chaos.

- Assigning specialised roles for exploration, implementation, testing, security review, and adversarial verification.

- Selecting appropriate models and reasoning levels for different types of work.

- Creating reusable skills, instructions, context systems, and domain-specific workflows.

- Connecting agents to external tools through secure, maintainable interfaces.

- Using permissions, persistent context, scheduled automations, and long-running workflows where they improve delivery.

- Managing context limits, agent handoffs, interrupted work, conflicting changes, and integration dependencies.

- Reviewing actual changes, test evidence, runtime behaviour, and migration safety instead of trusting an agent’s completion report.

- Establishing human approval gates around production deployment, data changes, security-sensitive code, and compliance decisions.

- Measuring agent quality, cost, latency, failure modes, and throughput.

- Knowing when parallel execution helps—and when work must remain sequential.

- Replacing tools when better options emerge without destabilising the engineering process.

- Building custom orchestration where available harnesses do not meet FrontierU’s needs.

Your results matter more than your preferred tooling. We want someone who understands the agent ecosystem deeply enough to make independent technical choices and explain them clearly.

Technical foundation

FrontierU is a modern full-stack web application backed by a relational database and external services for AI, payments, communications, and file storage. It has extensive automated testing and a continuous integration and delivery process.

You should be comfortable taking ownership of:

- Modern TypeScript-based web applications

- Component-based frontend and server-side development

- SQL and relational database design

- Schema changes and production migrations

- Cloud or serverless infrastructure

- Automated testing and continuous delivery

- Observability, backups, recovery, and incident response

- Secure integrations with external service providers

What we are looking for

You should have:

- A record of personally building and operating substantial production web applications.

- CTO, founding engineer, staff engineer, or equivalent end-to-end technical ownership experience.

- Strong modern web-development, server-side engineering, SQL, and relational-database expertise.

- Deep understanding of API design, authentication, authorization, transactions, concurrency, database migrations, queues, idempotency, and failure recovery.

- Production experience with cloud infrastructure, CI/CD, object storage, communications, payments, observability, backups, and incident response.

- Experience building products around language models, including structured outputs, prompt-injection defence, retries, model failures, evaluations, cost control, and safe degradation.

- Advanced, demonstrable experience orchestrating AI agents through one or more modern agentic environments.

- The ability to assess new agent tools independently rather than following one vendor’s preferred workflow.

- The ability to diagnose subtle failures that generated code and ordinary test suites may miss.

- Strong security and privacy judgment, particularly around GDPR, sensitive data, access control, auditability, and products used by minors.

- The ability to communicate technical risks and trade-offs clearly to a non-technical founder.

- The judgment to distinguish a genuine blocker from something that can be solved autonomously.

- The ability to hire, assess, and lead engineers when a human team becomes necessary.

A computer science degree is not required. Evidence of technical depth, ownership, judgment, and shipped work matters more.

Useful additional experience

Experience in any of the following would be valuable:

- Private communities, social products, marketplaces, education, or talent platforms

- Trust and safety, moderation, or safeguarding

- GDPR-sensitive or regulated products

- AI interviews, assessments, mentoring, or recommendation systems

- Payment systems and other sensitive third-party integrations

- Accessibility and performance engineering

- Multi-agent evaluation, reusable agent skills, tool integrations, or custom orchestration infrastructure

What we offer

- A full-time role from day one

- Remote work from anywhere, salary premium during the times you stay with us in Florence (digital nomad visa will be arranged)

- Competitive salary plus meaningful equity

- Direct influence over the product, architecture, and company

- Freedom to choose and evolve the agentic toolchain

- Authority to establish the engineering system and selectively build a team

- The opportunity to create an unusually capable agent-first technology organisation

How to apply

Please send:

- Your CV, profile, or relevant work history

- Examples of production systems you have personally owned

- A short description of an agent-driven engineering system you have set up

- The tools and orchestration patterns you chose—and why

- One example of an important mistake, security issue, or false completion claim you caught in an agent’s work

- A brief outline of how you would approach FrontierU’s first 30 days

We are looking for someone who can make agents move quickly without allowing the system to become careless.