AI-based IP research leads to more efficiency, less error rates: Manish Sinha, Founder-CTO, PatSeer

Artificial intelligence (Al) is increasingly driving innovation across several domains, including intellectual property (IP) research where it is improving search and process efficiency. In a conversation with TechCircle, Manish Sinha, Founder and Chief Technology Officer (CTO) PatSeer, an Indian AI-based business-to-business (B2B) software-as-a-service (SaaS) platform for IP research and intelligence discusses the advantages of incorporating AI into IP research and how it can benefit various industries. He also delves into semantic search (a method that helps find data based on the intent and contextual meaning of a search query) in IP research. Edited excerpts: 

What benefits does AI bring to IP research and intelligence?  

AI-powered tools streamline the research process by automating tasks like patent searches, document classification, and data extraction, thereby increasing research productivity and saving time. Manual classification, as we know, are prone to human errors. When replaced by trained AI classifiers that automatically learn from existing taxonomies, it ensures accurate predictions for newer, uncategorised records. These tools also help overcome language barriers, offering improved technology translation and predictive analytics forecasts on IP trends, litigation risks, and market opportunities, facilitating long-term IP strategy development. More importantly, AI-based text analysis tools standardise and clean patent data, ensuring researchers work with accurate and reliable information. 

How are machine learning (ML) and AI capabilities utilised in IP research? 

AI/ML is changing the way patent professionals and researchers extract insights from vast volumes of patent data. AI Classifiers utilise algorithms to categorise unstructured data, such as patent text, into sets of classes and create trained models for prediction. These models find application in various patent activities, including predictions on emerging trends, potential competitors, and innovation opportunities. Furthermore, generative AI algorithms like GPT-4, Claude, Llama, and others contribute to patent drafting and generate intelligent summaries. Additionally, they power inbuilt AI assistants through chat interfaces, enhancing the overall efficiency of IP research processes.  

How does AI-driven IP benefit various industries?  

To begin with, it optimises the patent research process, leading to significant time and resource savings. Moreover, it fortifies competitive advantages by furnishing insights into market trends and competitor activities. This, in turn, empowers strategic decision-making, assisting organisations in pinpointing potential partnerships, licensing opportunities, and areas ripe for innovation. Beyond the corporate sector, AI-driven IP intelligence extends its support to non-profit entities. It facilitates cost-effective patent searches, fostering innovation, particularly in critical domains such as healthcare and clean energy. It also enhances overall productivity, elevates capabilities, and contributes to a reduction in error rates. 

Are there industry-specific use cases that illustrate the benefits of AI-driven IP intelligence?  

There are notable instances of industry-specific use cases that underscore the tailored benefits of AI in patent research and intelligence. In the pharmaceutical sector, AI-powered IP intelligence proves instrumental in accelerating drug discovery processes. Within the high-tech sector, where Standard Essential Patents (SEPs) play a pivotal role, AI tools perform a standard essentiality check on patents. In the fashion industry, AI-based image recognition algorithms take centre stage. These algorithms are deployed to effectively search and monitor design patents, ensuring the safeguarding of original designs and proactively preventing potential infringements. There are many more use cases across sectors. 

In terms of IT spending, how is your company’s tech budget growing this year? Is that percentage higher, lower, or the same as the last few years?  

This year, our company is experiencing a growth in IT spending, with a 40% increase in investments specifically directed towards enhancing our IT infrastructure and expanding our online cloud services.  

How is the tech team being boosted in terms of the number of people and capabilities in the current year?  

Our technology team is undergoing significant expansion, particularly in the areas of DevOps, data science, and AI/ML roles. We anticipate a 20% increase in the overall size of the technical team by the end of this year, accompanied by a broadening of capabilities.  

On a scale of 1 to 10 where 10 is high and 1 is low, where would you put your organisation in terms of the digital transformation journey, and what is the short- to medium-term target on that front? 

We actively incorporate technology across various facets of our business operations. As a B2B SaaS product company, software forms the cornerstone of our product offerings. Our utilisation of multiple web-based tools spans customer relationship management (CRM), HRMS, IT security, marketing automation, and accounting. Given our extensive integration of technology, we would assess our organisation at a level 7 on the digital transformation scale. Looking ahead, our immediate objective is to integrate AI not only into our product but also into diverse applications employed by our development, sales, and marketing teams. 

How is semantic search changing the landscape of IP research? 

Semantic search is changing IP research by moving away from traditional keyword-based searches that rely on exact matches, and instead delving deeper to understand the context and meaning of words. This allows for more accurate and relevant results, making it easier for researchers to find the patents, and identify records that intentionally omit important keywords to hide patents from public discovery. By understanding the context, semantic search can uncover these hidden patents and provide a more comprehensive view of the intellectual property landscape. It also reduces the workload on IP and legal teams, who no longer have to spend extensive time and resources assessing the novelty of each idea presented. That said, it streamlines the patent search process, saving time and resources while ensuring comprehensive and precise results.