About the session
Ahmed Bashir and Ritesh Goru discuss the context engineering behind a more useful AI teammate, from system prompts and skills to tool guidance, hooks, and model selection.
They connect those design choices with benchmarking and real-world signals: ways to test changes, learn from failures, and improve an agent over time.
Key takeaways
- Identify the context and guidance choices that shape an agent's responses.
- Examine how benchmarking can test changes to the AI experience.
- Use real-world signals and failures to inform iteration.
Agenda
Context engineering for AI teammates
Prompts, skills, tools, hooks, and model selection
Benchmarking and real-world feedback
Speakers
Ahmed Bashir
Chief Technology Officer, DevRev
- RG
Ritesh Goru
