GridSynapse: Compare GPU Options Before You Spend
Elijah Paul
GridSynapse helps infrastructure teams compare GPU options before committing budget.
Compute decisions involve more than price. Teams need to weigh region, carbon, modeled capacity, modeled network lag, workload timing, and other hard constraints without losing the reason behind a recommendation.
I set the product direction and built the decision flow, Next.js and TypeScript interface, Python and FastAPI backend, optimization logic, and data model.
The workflow:
• Define the workload and hard constraints.
• Compare candidate compute options from a catalog.
• Apply cost, region, carbon, modeled capacity, and timing rules.
• Review the ranked recommendation and its reason trail.
The public product uses synthetic reference data for planning. It does not claim live provider capacity, reservations, or procurement.
Modeled comparison view with cost, region, carbon, and workload limits kept visible.
Planning workflow from workload constraints to a human-reviewed shortlist.
The result is a reviewable planning workflow for comparing options before a team spends money.
Like this project
Posted Jul 30, 2026
Built a compute planning console that compares catalog options by cost, region, carbon, modeled capacity, modeled network lag, and workload rules.
