Proposal: Canton Fee Estimator — Pre-Submission Transaction Cost Intelligence

Development Fund Proposal Submission

Proposal file: /proposals/canton-fee-estimator.md

---

Summary

Canton Fee Estimator is an open-source SDK and API that gives Canton developers accurate pre-submission transaction cost estimates — the eth_estimateGas equivalent Canton is missing today. Every funded DeFi project (ACME, Alpend, Cantex, Hecto) currently hard-codes its own fee buffer. When Canton upgrades its fee model, they all break independently. This tool fixes that with one maintained open-source standard.

The repository at jalajnagar88-stack/canton-fee-estimator is active — it includes a working Scan API integration that has collected real fee data from Canton TestNet, a TypeScript SDK skeleton, FastAPI backend, and documented findings in LEARNINGS.md. This proposal requests 340,000 CC across 3 milestones to complete, harden, and maintain it as a permanent public good.

Estimation approach: historical percentile distribution from the public Scan API (p50/p75/p95), with attribute-based modelling as fallback. No privileged API access required — works with Canton's existing public surface today.

---

Checklist

• [x] Proposal file added under /proposals/
• [x] Milestones and funding amounts defined
• [x] Acceptance criteria included
• [x] Alignment with Canton priorities described

---

Notes for Reviewers

I have been building on Canton TestNet and studying the Scan API fee structure prior to this proposal. Key observation documented in LEARNINGS.md: the Ledger API v2 does not expose a dry-run simulation endpoint, so the estimation engine uses historical percentile modelling from Scan data — which is sufficient for ≥90% of developer use cases and more practically deployable today.

Seeking a Tech & Ops Committee champion. Happy to demo the TestNet Scan API integration live, answer questions, or adjust scope based on committee feedback.