# Aviro

> Post-training datasets and reinforcement learning environments for AI models.

Aviro designs challenging worlds where AI models learn to explore and solve problems. These worlds become post-training datasets and reinforcement learning environments for training the next frontier of AI. The work spans coding, computer use, knowledge work, and research.

For partnerships, research, or enterprise inquiries, contact founders@aviro.ai.

## Company

- [Company overview](https://www.aviro.ai/index.md): What Aviro builds and its four areas of work, in Markdown.
- [Homepage](https://www.aviro.ai/): The current company introduction and interactive illustrations.
- [Book a demo](https://cal.com/aviro/demo): Discuss a project or partnership with the team.

## Benchmarks and research

- [Benchmarks](https://www.aviro.ai/benchmarks): Public benchmarks and evaluation results.
- [C4 benchmark](https://www.aviro.ai/benchmarks/c4): Grounded reasoning across enterprise documents and workflows.
- [Introducing C4 and Ebla-1](https://www.aviro.ai/writeups/c4-ebla): Research introducing the benchmark and model.
- [Context engineering](https://www.aviro.ai/writeups/context-engineering): Research on context engineering for long-horizon agents.
- [Writeups](https://www.aviro.ai/writeups): Published research and technical articles.

## Optional

- [Careers](https://www.ycombinator.com/companies/aviro/jobs): Open roles at Aviro.
- [Privacy Policy](https://www.aviro.ai/privacy-policy): Information collection, use, and privacy practices.
- [Terms of Service](https://www.aviro.ai/terms-of-service): Terms governing use of Aviro services.
- [Cookie Policy](https://www.aviro.ai/cookie-policy): Cookies and similar technologies.
- [Service Level Agreement](https://www.aviro.ai/sla): Service commitments.