#70
Monthly Rank
#197
Lifetime Rank
Share
8 September 2026
T

Trademark OS

An intelligence platform for trademark professionals. AI-driven assistance from managing clients to generating documents.

Gallery

About the Project

TrademarkOS is a practice-management platform built specifically for Indian trademark law. Indian practice runs on the Trade Marks Act 1999 / Rules 2017 and is uniquely deadline-hostile: missing the 30-day examination-response window, the 4-month opposition window, or the 2-month counter-statement window causes irreversible loss of rights "by operation of law, not merits." Most firms still track this on spreadsheets.

The core idea is a two-axis status model that separates what the Registry has actually done (Registry Status) from what the firm believes is happening (Firm Phase), with an automated Divergence Check that surfaces the gap in real time, catching exactly the failure mode that loses marks by operation of law. Contentious work is modeled the way Indian practice actually works: a typed Proceeding (opposition, rectification, appeal) sits as a child of a continuing Matter, because a single mark crosses between prosecution and litigation repeatedly, sometimes sitting on both sides at once.

On top of that: an AI Pre-Filing Clearance Report with a Self-Portfolio Collision Check, screening a proposed mark against the same client's own portfolio, not just third parties; Mirror Mode AI drafting for trademark applications and cease-and-desist notices, styled to a firm's own documents; a deadline-rules engine covering the Act's statutory timelines; a 6-band permission matrix (Partner/Attorney/TM Agent/Associate/Paralegal/Front Desk) with field-locking after filing; a Client entity with GST-aware financials and a Relationship Health Score deliberately distinct from deadline urgency; and interactive Knowledge Base guides whose worked examples call the platform's real formulas, backed by build guards that fail CI the moment a formula and its documentation drift apart.

It's one Next.js/TypeScript app end-to-end, real Prisma reads/writes throughout, covering leads through litigation: net-new lead intake with live WIPO Nice Classification search across terms, Clients, Matters, Contacts, Tasks, Reports, notes with @mentions and a unified timeline, platform-wide folders with real file storage, and a full Settings/Resources suite.

Practice Areas

Key Features

  • Two-axis Registry Status / Firm Phase model with automated Divergence Check — catches the exact gap that loses marks "by operation of law"
  • Litigation/prosecution split: typed Proceedings (opposition, rectification, appeal) as children of a continuing Matter, matching how a mark actually moves through Indian practice
  • AI Pre-Filing Clearance Report with a Self-Portfolio Collision Check (screens against the same client's own portfolio, not just third parties)
  • Mirror Mode AI drafting for Trademark Applications and Cease-and-Desist notices, styled to the firm's own documents
  • Deadline-rules engine covering Trade Marks Act 1999 / Rules 2017 statutory timelines, with anchor-based rules so one status can arm multiple clocks
  • 6-band permission matrix (Partner/Attorney/TM Agent/Associate/Paralegal/Front Desk)
  • Leads rebuilt as net-new-only: conversion produces a Client (not a Matter), backed by live WIPO Nice Classification search across 10,123 terms
  • Client entity with GST-aware financials rollup and a Relationship Health Score deliberately distinct from deadline urgency
  • Platform-wide folders with real file upload/download across a 15-folder TM-prosecution taxonomy
  • Polymorphic notes with @mentions, per-tab tagging, a unified timeline, and notification-latency analytics
  • Interactive Knowledge Base guides that call the platform's real formulas (not restate them), backed by 5 build guards that fail CI if a formula or contract drifts from its docs
  • Archive subsystem, an editable-assumptions Valuation Estimator, and a document-derived timeline

About the Creator

NG
Nikhil Goyal
Product Operations Specialist at Evenup
LinkedIn