Frequently Asked Questions
Everything you need to know about working with AI agent engineering teams.
Working With Us
What exactly is an "AI agent team"?
An AI agent team is a squad of specialized AI agents — each focused on a specific area like frontend development, backend engineering, testing, or DevOps — working together under the supervision of senior human engineers. Think of it as an engineering team where AI handles the execution and humans handle the strategy and quality control.
How is this different from using ChatGPT or Copilot?
Tools like ChatGPT and Copilot assist individual developers with code snippets. Our agent teams are autonomous — they pick up tasks, write complete features, create pull requests, run tests, and deploy code. They work independently 24/7 on your codebase, not just when a developer prompts them. Plus, every significant change is reviewed by a senior human engineer.
How quickly can you start on my project?
Most projects kick off within 1-2 weeks of signing. We spend the first few days on discovery and architecture with our human engineers, then deploy your agent squad. You'll typically see the first pull requests within 48 hours of squad deployment.
Do I need to provide technical specs?
Not necessarily. If you have detailed specs, great — our agents can work from them directly. If you have a rough idea, our senior engineers will help you define the technical approach during the discovery phase. We meet you where you are.
Can your agents work with my existing codebase?
Yes. Our agents integrate with your existing Git repositories, CI/CD pipelines, and project management tools. They learn your code conventions, follow your branching strategy, and submit PRs just like any team member would.
Technology & Security
Is my code safe?
Absolutely. Your code stays in your repositories. We access it through standard Git protocols with the same security controls you'd give any engineering team member. We sign NDAs, follow SOC 2 practices, and can work within your specific security requirements.
What tech stack do you support?
Our agents work across most modern stacks — TypeScript/JavaScript (React, Next.js, Node.js, Vue), Python (Django, FastAPI, Flask), Go, Rust, and more. We're strongest in web and cloud-native development but can handle mobile (React Native, Flutter) and infrastructure (Terraform, Kubernetes) as well.
Do you use open-source or proprietary AI models?
We use a combination of leading AI models and our own fine-tuned systems. The specific model mix is optimized per project — some tasks need deep reasoning, others need speed. Our human engineers continuously tune the agent configurations for your specific codebase.
What happens if an agent writes bad code?
Every code change goes through automated quality gates — linting, type checking, automated tests, and security scanning. Significant changes also get human review. Bad code gets caught and corrected before it reaches your main branch. Our agents also learn from corrections, improving over time on your specific project.
Process & Quality
How do you ensure code quality?
Multiple layers: automated test suites (agents write and maintain tests), static analysis, code review by AI agents specialized in review, and human engineer oversight on critical paths. We maintain test coverage targets and quality metrics that we share with you.
How do I communicate with the team?
You'll have a dedicated Slack or Discord channel with your senior engineer and visibility into agent activity. You can also communicate through your project management tool (Linear, Jira, GitHub Issues). Need a feature? Create a ticket and agents will pick it up.
What if I'm not happy with the work?
We believe in transparency and iteration. You have full visibility into every commit and PR. If something isn't right, flag it and we'll correct course immediately. Our human engineers are there to ensure alignment with your vision.
Do agents work on one project at a time?
Your agent squad is dedicated to your project. They maintain context about your codebase, your conventions, and your goals. This isn't a shared pool — it's your team.
Getting Started
What does a typical engagement look like?
Most engagements start with a discovery call, followed by 2-3 days of architecture and setup, then continuous development. Minimum engagement is typically 1 month. Most clients continue for 3-12+ months as they see results.
What size projects do you take on?
From MVPs for solo founders to major platform builds for funded startups and enterprise teams. If it involves writing software, our agent teams can help. We'll be honest if a project isn't a good fit for our model.
Do I own the code?
Yes, 100%. All code produced is yours. It lives in your repositories, under your accounts. We don't retain any rights or copies after an engagement ends.