Context
As lead developer, I built a system of intelligent agents to automate tasks in a business environment, applying generative AI techniques and integrating with external services.
Approach
- AI agents. LLM-based components for automation and information processing.
- Integration. Connections to external services through APIs, and the processing logic behind them.
- Architecture. Reusable, configurable components built with Python and LangChain.
- Deployment. Services and automated tasks running in Kubernetes environments.
- Observability. Mechanisms for tracing, monitoring and evaluating the runs.
Outcome
I led the development of a modular, maintainable solution for automating business processes with AI agents, including tools to monitor how it works and make it easier to evolve.