Optimal: Cost effective infra with agentic inbox
Hackathon-stage; needs real validation that repo analysis beats manual estimation.

Yet another cloud cost optimizer competing with Vantage, CloudHealth, and Azure Cost Management.
DevOps engineers and cloud architects managing Azure spending
Vantage · CloudHealth · Azure Cost Management
The frontend is a React/TypeScript that shows AI generated cost recommendations, rightsizing opportunities, and idle resources into a prioritized dashboard. The backend is a Java Spring Boot API using Semantic Kernel to connect to your Azure environment via a read-only allowlist of Azure tools (Resource Graph, Cost Management, Monitor). It pulls live information and feeds it into an AI agent. The agent analyzes your actual resource utilization and expenditure to generate prioritized cost optimization suggestions. We enforce strict read-only access at the backend layer, ensuring the AI can never perform any write or delete operations within your environment.
We just launched our MVP and are looking for early users to test our product and provide feedback. We are offering it completely free in exchange for your honest insights.
Currently, we require you to reach out to the email on our website to get an account set up. We’ve worked to make this process as painless as possible, with a quick 10-minute setup on our end. You can reach out to me specifically at [email protected].
Hackathon-stage; needs real validation that repo analysis beats manual estimation.
Title says code generation, screenshot shows Azure cost tool — something's wrong here.
Offline cost estimation for Terraform that beats Infracost on speed and privacy.
Beats PyTorch eager by 5.29x on RMSNorm using autonomous agent loops.
Autonomous kernel optimizer that won MLSys contest with 34.93x speedup.
Pinpoints exact cost drivers like expired Reserved Instances instead of just showing service-level charts.