About the role
As an AI Agent Engineer at TheWebVale, you will build autonomous multi-step agent systems for client platforms. You will create agents that conduct market research, reconcile booking records, automate client intake, and coordinate complex multi-service tasks for business clients.
You will work in close coordination with backend engineers and our founder throughout our two-to-three week release cadences. You will design agent workflows with deterministic safeguards, error recovery loops, and clear human approval checkpoints.
What you will do
- Architect autonomous agent workflows that execute multi-step business operations.
- Implement structured tool-calling interfaces connecting agents to databases and external APIs.
- Design loop detection and error recovery mechanisms to prevent run-away agent execution.
- Build state persistence layers storing agent context, conversation history, and steps.
- Create human-in-the-loop review screens where agents request confirmation before actions.
- Conduct automated stress tests simulating unpredictable tool responses and network drops.
What you will need
- Possess 3–5 years of software engineering experience with focused AI agent development.
- Demonstrate advanced proficiency in Python or TypeScript with async system design.
- Show practical experience building agentic workflows using LangGraph, CrewAI, or custom loops.
- Understand function calling protocols, schema validation, and context management.
- Possess strong systems debugging skills when tracing non-deterministic execution paths.
- Communicate agent capabilities, safety limits, and execution risks with transparency.
Nice to have
- Experience integrating browser automation tools like Playwright into agent toolsets.
- Familiarity with vector memory systems and long-term agent state storage.
- Knowledge of sandboxed execution environments for running agent-generated scripts.