AI infrastructure and model selection
ALTSELF
ALTSELF is Om Batavia’s AI model routing project, selecting models by task complexity, quality, latency, and cost to reduce unnecessary AI spend.
Built by Om Batavia, a student at The Riverside School, Ahmedabad.
Problem
Using an expensive AI model for every request can add unnecessary cost and latency. Simple tasks and complex tasks have different requirements, so a single model choice may not be the right fit for every part of an AI product.
Solution
ALTSELF is a routing layer that chooses an AI model based on task complexity, expected output quality, latency, and cost. Its model selection and orchestration logic are designed to match the task with an appropriate model while keeping those trade-offs explicit.
Context
The project explores AI cost optimization and practical infrastructure for AI applications. It is designed for teams deciding how to balance response quality, speed, and operating cost across different tasks, with the aim of avoiding unnecessary model spend.
Role
Founder — AI orchestration logic and product direction. Om Batavia is building ALTSELF around the decisions involved in model routing: what a task needs, how quickly it needs a response, and what that response should cost.
Build something useful.
Have a related AI workflow, product idea, or operational problem? Share the context and desired outcome.
Contact Om