How Softweb Solutions built a patent data intelligence system powered by Needle, GenAI framework, for DorothyAI

NLP-powered REST API for large-scale patent data processing

Our team built a REST API infrastructure to handle DorothyAI’s global patent dataset. The API uses Needle’s natural language processing layer to interpret user queries and translate them into optimized data retrieval logic. Users receive contextually relevant patent information without writing Boolean search strings or understanding underlying database schemas.

The API also supports multilingual input, allowing professionals to query the system in their native language. Needle’s NLP capabilities manage language parsing and semantic alignment before executing the search. This broadens the usability of DorothyAI’s platform for an international user base.

Real-time patent data visualization and trend analytics

We integrated real-time data visualization tools into DorothyAI’s chat interface. When users query patent trends, competitive activity, or filing volumes, the system generates charts and graphs within the same window. Visual outputs are rendered dynamically based on the query. It gives users an immediate and intuitive picture of the patent landscape.

The visualization layer is connected to live patent data, which means charts reflect up-to-date information. Legal teams and R&D professionals can spot emerging trends, identify competitor patent activity, and assess market positioning immediately. This approach shifts patent analysis from a time-intensive reporting task to a real-time, interactive workflow.

AI-powered document upload and patent comparison engine

Users can upload scientific documents, PDFs or Word files, into the platform. Needle’s AI summarizes and compares them against existing patents in seconds. The comparison engine identifies structural overlaps, claim similarities, and key differentiators. Results are presented in a concise, readable format suited for rapid decision-making.

Researchers submit invention disclosures and product data sheets or academic papers to receive actionable insights. Patent attorneys and R&D leads use this feature to assess novelty risk while preparing for prosecution and guiding filing strategy with speed and confidence.

Multi-language support for a global patent research audience

We extended DorothyAI’s platform to support multilingual query processing. It enables users across geographies to interact in their preferred language. Needle’s NLP layer handles semantic understanding across languages before the query reaches the search and retrieval engine. This preserves query intent and ensures results are as precise in French, German, or Japanese as they are in English.

The multi-language layer also improves accessibility for international IP teams that collaborate. Users no longer need to translate queries manually or rely on approximations. The system processes the query natively and returns results that align with the user’s original intent. This feature brings DorothyAI’s patent intelligence to a larger global audience.