Qodo (formerly Codium) at GTC
See Agentic AI for testing reviewing and writing code, see how test-driven code generation improves the quality and accuracy of your code. Come by our booth #3210 to see a live demo.
We are excited to showcase our multi-agent platform to review, test and write code.
- Automates code review workflows and
improves code quality - Generate meaningful tests and increase
your coverage - Generate code according to your
company best practices
Set a demo at the booth and collect
your personal swag
New model
Best code embedding model performance
- Qodo-Embed-1-1.5B outperforms OpenAI and Salesforce models in code embedding while being significantly smaller (1.5B vs. 7B parameters), setting a new benchmark for efficiency in code understanding.
- Optimized for low-cost GPUs, it enables advanced code search and retrieval, making powerful AI-driven development tools accessible to teams of all sizes.
Trained on
Talk
Real-world coding agents starts with a good code embedding model
High-performance code embedding is a critical building block, and it needs to be designed and trained to work well for the desired workflowsIn this talk, we will share technical details on how at Qodo we trained code embedding and how we use it in Qodo’s agents.
Not going to make it this year?
Join our monthly demo webinar to see testing-driven code generation in action