Back to projects

Case Study

Slate

Interactive patent collaboration platform

PHPCakePHPApache SolrOCRNLPMySQL

Role

Architect & Team Lead

Outcome

Reduced client patent analysis time by 70%

Proprietary / internal project

Overview

Slate is a collaboration workspace built for patent analysts to review, annotate, and search thousands of patent documents in real time. It replaced a slow, manual document-review workflow with an interactive platform that ingests scanned filings, extracts text via OCR, and layers NLP-driven annotation and full-text search on top.

Problem

Patent analysts at enterprise clients like Adidas, Huawei, and a prominent Middle Eastern education and research foundation were manually reading and cross-referencing large volumes of scanned patent filings, a process that was slow, error-prone, and difficult to collaborate on across distributed teams.

Features

  • OCR pipeline to digitise and index scanned patent documents at scale
  • NLP-assisted annotation to surface relevant clauses and prior art
  • Apache Solr-powered full-text search across the entire document corpus
  • Real-time collaborative review workflow for distributed analyst teams

Key Learnings

  • Designing search relevance for highly technical, jargon-dense legal text requires custom tokenization and ranking beyond out-of-the-box Solr configs.
  • Collaborative annotation tools need careful conflict handling when multiple analysts work the same document simultaneously.