✅
Step 1 — Select your use-cases
Open 1. Select Use-cases and pick a package first — nothing is listed until you do. Narrow further by category or search, tick what the site needs, set the camera count for that selection and hit Add to estimator →. A 🛠 marks a use-case that runs on-cloud today and needs development work for on-prem; its cost and timeline are shown alongside.
🖥️
Step 2 — Estimate the load
Your picks land in 2. Estimate Load, where each row’s camera count can still be changed. Fill in Total cameras only when different use-cases watch different cameras. Choose Cost-Optimal (VLMs excluded) or Accuracy Optimal (full model stack), then click Estimate Load for the vCPU, GPU VRAM, GPU cores and RAM the deployment needs — with a per-use-case breakdown and a what-if scaling table underneath.
🔍
Step 3 — Find your server
Go to 3. Find Server. Rank the server library by cost or by server count, or enter a custom spec to learn how many boxes of it you would need. The split modes decide whether use-cases and cameras may be spread across machines. Utilisation bars show the headroom left over, and every price follows the INR / USD switch in the top bar.
📁
Step 4 — Keep it and share it
Every estimate auto-saves to Export/Import — name it, compare two scenarios side by side, or export the history as a file to send on. Server library lists the hardware you can size against, and Data source holds the workbook all of this is computed from: anyone can download it, admins can publish an updated one for everyone.