Glassbox's mission is to empower enterprises to shape trusted, frictionless digital experiences.
Glassbox is a leading force in shaping digital experiences. It helps organizations uncover digital issues, boost conversion rates, enhance accessibility, prevent fraud, and more. Leveraging AI-driven customer intelligence, Glassbox enables enterprises to deliver secure, proactive, and preventative digital experiences. Its solutions are trusted by highly regulated organizations, including SoFi, Cal, and many others. We are growing and have been recognized by G2 as one of 2024's Top 50 Software Companies in the world.
We are building an AI-native R&D platform that transforms how engineering teams operate.
Our vision is to move from reactive workflows to intelligence-driven systems, where AI agents are embedded across the entire software development lifecycle - from code to production.
These agents will both:
- Work alongside engineers as copilots
- Operate independently as autonomous systems
We are looking for a hands-on AI Agent Engineer to design and build production-grade AI systems that integrate deeply into engineering workflows.
This is not a research role and not a prompt-engineering role.
You will build real systems that:
- Are used daily by engineers
- Run independently and take action
- Deliver measurable impact on speed, quality, and reliability
- Design and build AI agents that:
- Assist engineers in real-time (copilot mode)
- Operate autonomously and take actions (self-sufficient mode)
- Develop end-to-end agent workflows across systems such as:
- CI/CD pipelines
- Code repositories
- Observability and monitoring tools
- Support and ticketing systems
- Build core components:
- RAG pipelines (retrieval, embeddings, vector databases)
- Tool-integrated agents (APIs, services, workflows)
- Agent orchestration, memory, and state management
- Turn LLM capabilities into production-grade systems by:
- Handling failures, retries, and edge cases
- Managing latency, cost, and scalability
- Implementing observability (e.g., Langfuse or similar)
- Work closely with engineering teams to ensure:
- Real adoption
- Seamless integration into workflows
- Balance between autonomy and human control
- Strong backend or platform engineering experience (2–5+ years)
- Proven experience building and deploying LLM-based agents in production
- Hands-on experience with:
- LangChain, LangGraph, or similar frameworks
- LLM APIs (OpenAI, Anthropic, etc.)
- RAG systems (vector DBs, embeddings, retrieval pipelines)
- Tool-based agents (APIs, workflows, automation)
- Experience with Claude (Anthropic), including Claude Code - mandatory
- Used as part of development workflows or integrated into systems
- Not limited to chat usage
- Experience with observability tools such as Langfuse (or similar)
- Strong coding skills in TypeScript or Python
- Experience building multi-agent or autonomous systems
- Experience designing human-in-the-loop workflows
- Background in DevOps, CI/CD, QA, or DevEx platforms
- Experience with event-driven systems (Kafka, queues)
- Experience integrating AI into:
- Code review
- Testing / QA
- Incident analysis
- A builder - you ship systems end-to-end
- You think in systems and workflows, not just prompts
- You are comfortable going from idea to production
- You care about impact, reliability, and real-world usage
You'll be building the foundation of an AI-native engineering platform, where:
- Agents are first-class components across R&D
- Systems are autonomous, connected, and continuously improving
- Engineering teams are augmented - not replaced - by AI
This is an opportunity to shape how modern engineering organizations operate in the AI-first era.
At Glassbox, we value curiosity, ownership, and a constant drive to learn and improve, no matter the gender, nationality, religion, or background. We believe diverse perspectives make us better, and that potential matters just as much as experience, and encourage people of all shapes and sizes to apply.
If this role excites you and you're motivated to make an impact - we'd love to hear from you, even if you don't meet every listed qualification.
