Welcome, we're extremely happy to have you join our community of GenAI application builders.
Arato helps technical and not-so-technical team members take AI products from early-stage idea to production-grade feature with tooling for experimentation, evaluation, deployment, CI/CD, monitoring, and collaboration.
Our documentation will guide you through key concepts as well as best practices in building GenAI applications and experimentation.
Here you'll find all the relevant guides for understanding key concepts from running experimentation to production-grade GenAI applications. If you are missing anything or need any firsthand guidance, please don't hesitate to contact us.
GenAI has changed how applications get built. It is stemmed in the fact that GenAI is non-deterministic, so it's impossible to determine whether what we are building will actually work. The only way to do so is to experiment repetitively and iterate until you achieve the desired results.
Key Challenges in GenAI Development
AI development is often slowed down by silos between technical and business teams, leading to:
Comprehending the nondeterministic nature of LLMs
Trying to fit GenAI development into existing software development processes
AI Applications shipped with basic testing and no structured process
Assumptions or manual evaluations of output
Insufficient monitoring of production use cases
Multiple stakeholders slowing down development process
Arato addresses these by providing a GenAI-native, unified platform with the tools to build, experiment, deploy, monitor, and optimize your GenAI projects in a systematic manner.
Arato provides several key functionalities under one integrated platform:
Low/No-code environment: Simplified interface to run tests & experiments
Experimentation at scale: Easy to alternate and iterate
AI Evaluation framework: Out-of-the-box and custom Evals to provide insights on experiment results
Version control: Track prompt and model configuration changes and their effect on the experimentation results
Collaboration: Shared workspace with your teams
Monitoring: Track application behavior in production
At the heart of Arato, we have simplified the ability to build and run experiments, iterate, and optimize at scale to achieve the desired outcomes. For that purpose, we have created the Arato Notebook, a powerful framework that brings together all the key concepts of AI experimentation into a streamlined experience.