Meet our guest Leon Lobo, Software Engineering Director, Oracle.
Startups offering AI-Assisted Coding tools are the fastest-growing companies ever….We are now taking a deep dive in the podcast to explore how this technology can be used for enterprise environments with large codebases and high requirements for security, compliance, and confidentiality.
In this episode we talked about:
- Enterprise AI testing requires new approaches - Moving from deterministic to probabilistic testing using vector similarity while preventing functional regressions that could crash critical business systems.
- RAG enables secure AI adoption without training models - Keep customer data secure and stay current with latest AI capabilities by feeding documentation through RAG pipelines instead of model training.
- Strong architecture becomes critical with AI coding tools - AI boosts productivity for documentation and routine tasks, but enterprise applications still need experienced architects to prevent maintenance nightmares.
- Oracle uses Cline and other modern AI coding tools - Practical implementation of repository-based development tools combined with guardrail services to catch security vulnerabilities and competitor code.
- Hackathons evolve toward product-focused collaboration - Product managers can now build AI agents directly while engineers focus more on business problems than technical implementation details.
Enjoy!
